{"id":"3cbe1690-b50d-4132-83ae-7d0cc4981092","entityType":"agent","slug":"clawhub-deciqai-cynefin","name":"Cynefin","canonicalUrl":"https://www.xpersona.co/agent/clawhub-deciqai-cynefin","canonicalPath":"/agent/clawhub-deciqai-cynefin","generatedAt":"2026-10-11T03:56:19.051Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-11T01:37:00.444Z","emptyReason":null},"description":"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know w... Skill: Cynefin Owner: deciqai Summary: Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know w... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T17:56:40.866Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/cynefin.json) v1.0.4 | 2026-07-09T11:16:45.631Z | user Refresh: 20","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 1.2K downloads reported by the source. Last updated 10/11/2026.","installCommand":"clawhub skill install s17a4mqcnk515kvaca5ze55d0x88pfpx:cynefin","sourceUrl":"https://clawhub.ai/deciqai/cynefin","homepage":"https://clawhub.ai/deciqai/skills/cynefin","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/deciqai/cynefin","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/deciqai/skills/cynefin","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":62,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know w..."},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-11T01:37:00.444Z","emptyReason":null},"protocols":[{"protocol":"OPENCLEW","label":"OpenClaw","status":"self-declared","notes":"Declared in the public agent profile."}],"capabilities":[],"verifiedCount":0,"selfDeclaredCount":1,"capabilityMatrix":{"rows":[{"key":"OPENCLEW","type":"protocol","support":"unknown","confidenceSource":"profile","notes":"Listed on profile"}],"flattenedTokens":"protocol:OPENCLEW|unknown|profile"}},"adoption":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T01:37:00.444Z","emptyReason":null},"stars":null,"forks":null,"downloads":1203,"packageName":null,"latestVersion":"1.0.5","tractionLabel":"1.2K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T01:37:00.380Z","emptyReason":null},"lastUpdatedAt":"2026-10-11T01:37:00.444Z","lastCrawledAt":"2026-10-11T01:37:00.380Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-12T01:37:00.380Z","lastVerifiedAt":null,"highlights":[{"version":"1.0.5","createdAt":"2026-07-16T17:56:40.866Z","changelog":"Description tail link + agents machine-readable metadata line (deciqai.com/s/cynefin.json)","fileCount":7,"zipByteSize":17635},{"version":"1.0.4","createdAt":"2026-07-09T11:16:45.631Z","changelog":"Refresh: 2024-2026 AI-era worked examples added (strategy/leadership + systems/game-theory batch)","fileCount":7,"zipByteSize":17758},{"version":"1.0.3","createdAt":"2026-07-08T10:58:35.880Z","changelog":"Footer now uses /c/ short link (fixes UTM truncation when SKILL.md is read in a terminal)","fileCount":6,"zipByteSize":13125},{"version":"1.0.2","createdAt":"2026-07-08T00:43:29.472Z","changelog":"Refreshed content + GitHub star link in footer","fileCount":6,"zipByteSize":13061},{"version":"1.0.1","createdAt":"2026-07-07T20:31:55.489Z","changelog":"Add catalog categories and topics","fileCount":5,"zipByteSize":9719},{"version":"1.0.0","createdAt":"2026-06-27T09:17:12.895Z","changelog":"Initial publish","fileCount":5,"zipByteSize":9881}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s17a4mqcnk515kvaca5ze55d0x88pfpx:cynefin","setupComplexity":"low","setupSteps":["Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.","Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data."],"contract":{"contractStatus":"missing","authModes":[],"requires":[],"forbidden":[],"supportsMcp":false,"supportsA2a":false,"supportsStreaming":false,"inputSchemaRef":null,"outputSchemaRef":null,"dataRegion":null,"contractUpdatedAt":null,"sourceUpdatedAt":null,"freshnessSeconds":null},"invocationGuide":{"preferredApi":{"snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-11T03:56:19.047Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-cynefin/trust"}},"reliability":{"evidence":{"source":"runtime-metrics","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No trust, reliability, or runtime telemetry is available."},"trust":{"status":"unavailable","handshakeStatus":"UNKNOWN","verificationFreshnessHours":null,"reputationScore":null,"p95LatencyMs":null,"successRate30d":null,"fallbackRate":null,"attempts30d":null,"trustUpdatedAt":null,"trustConfidence":"unknown","sourceUpdatedAt":null,"freshnessSeconds":null},"decisionGuardrails":{"doNotUseIf":["Contract metadata is missing or unavailable for deterministic execution."],"safeUseWhen":[],"riskFlags":["missing_or_unavailable_contract","trust_data_unavailable","schema_references_missing"],"operationalConfidence":"low"},"executionMetrics":{"observedLatencyMsP50":null,"observedLatencyMsP95":null,"estimatedCostUsd":null,"uptime30d":null,"rateLimitRpm":null,"rateLimitBurst":null,"lastVerifiedAt":null,"verificationSource":null},"runtimeMetrics":{"successRate":null,"avgLatencyMs":null,"avgCostUsd":null,"hallucinationRate":null,"retryRate":null,"disputeRate":null,"p50Latency":null,"p95Latency":null,"lastUpdated":null}},"benchmarks":{"evidence":{"source":"no-benchmark-data","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No benchmark suites or observed failure patterns are available."},"suites":[],"failurePatterns":[]},"artifacts":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"high","updatedAt":"2026-10-11T01:37:00.444Z","emptyReason":null},"readme":"Skill: Cynefin\n\nOwner: deciqai\n\nSummary: Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know w...\n\nTags: latest:1.0.5\n\nVersion history:\n\nv1.0.5 | 2026-07-16T17:56:40.866Z | user\n\nDescription tail link + agents machine-readable metadata line (deciqai.com/s/cynefin.json)\n\nv1.0.4 | 2026-07-09T11:16:45.631Z | user\n\nRefresh: 2024-2026 AI-era worked examples added (strategy/leadership + systems/game-theory batch)\n\nv1.0.3 | 2026-07-08T10:58:35.880Z | user\n\nFooter now uses /c/ short link (fixes UTM truncation when SKILL.md is read in a terminal)\n\nv1.0.2 | 2026-07-08T00:43:29.472Z | user\n\nRefreshed content + GitHub star link in footer\n\nv1.0.1 | 2026-07-07T20:31:55.489Z | user\n\nAdd catalog categories and topics\n\nv1.0.0 | 2026-06-27T09:17:12.895Z | user\n\nInitial publish\n\nArchive index:\n\nArchive v1.0.5: 7 files, 17635 bytes\n\nFiles: examples/apollo-13-1970-mission-response.md (6142b), examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md (9098b), examples/sorting-ai-decisions-by-domain-2024-2026.md (8583b), references/sources.md (1921b), skill-card.md (2298b), SKILL.md (7544b), _meta.json (126b)\n\nFile v1.0.5:SKILL.md\n\n---\nname: cynefin\ndescription: \"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know what to do first', 'best practice doesn't apply here', 'we need a different approach'. Do NOT activate when: situation is unambiguously routine (execution only); a specialized tool (OODA, expected value) already fits. More: deciqai.com/c/cynefin\"\n---\n\n# Cynefin\n\n## Overview\n\n**Cynefin** (pronounced \"kuh-NEV-in\"; Welsh for \"habitat\") is a sense-making framework by Dave Snowden (IBM, 1999). Its claim: the right decision approach depends on which of five domains the situation falls into — **Clear** (obvious cause-effect, use SOP), **Complicated** (knowable with expertise, use analysis), **Complex** (emergent, probe first), **Chaotic** (absent cause-effect, act first), **Confused** (unknown domain, decompose first). The most common and costly error: treating Complex problems as Complicated.\n\nComposes with `ooda-loop`, `feedback-loops`, `antifragile`, `first-principles`.\n\n## When to Use\n\n- A familiar approach has stopped working and you can't articulate why\n- Experts disagree on the right answer — a crisis unfolding where the previous playbook doesn't apply\n- \"Best practices from X\" imported without checking if the domain matches\n- A team is over-planning something emergent, or \"let's get more data\" when data won't come without action\n- Allocating AI capex or racing AI-native competition: deciding which AI bets are engineering (Complicated), emergent agent/adoption experiments (Complex), or live incidents (Chaotic)\n\n**Not when:** domain is unambiguously Clear (execution only); small-stakes one-shot; specialized framework already fits.\n\n## Coaching Novices (Adaptive Front Door)\n\n**Engine mode:** concrete case → run The Process. **Coach mode:** unfamiliar → guide step by step.\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. Classify the situation (Clear/Complicated/Complex/Chaotic) and match decision method to domain.\n2. Check fit: if unambiguously routine (Clear), skip framework.\n3. Elicit their real case — decision, current method, cause-effect structure.\n> **[WAIT — do not advance until user responds]**\n4. Are cause-effect relationships obvious, knowable, emergent, or absent? Is current method matched?\n> **[WAIT — do not advance until user responds]**\n5. Close: named domain + matched decision method + boundary watch.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe:** `Decision/situation: | Current approach: | What worked/not: | Stakeholders:`\n\n### Step 2: Diagnose the domain\n```\nObvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)\n```\nDiagnostics: 5 experts converge? (Yes → Complicated; No → Complex). Standard best practice works? (Yes → Clear/Complicated; No → Complex/Chaotic). Interventions predictable? (Yes → Clear/Complicated; No → Complex/Chaotic).\n\n### Step 3: Match approach to domain\n```\nClear: Sense→Categorize→Respond (SOP/automate) | Complicated: Sense→Analyze→Respond (experts)\nComplex: Probe→Sense→Respond (safe-to-fail experiments, amplify wins)\nChaotic: Act→Sense→Respond (establish order, then re-classify) | Confused: decompose, classify each part\n```\n\n### Step 4: Check boundary movement + choose intervention\n```\nDomain shifted? (Complicated→Complex from disruption? Clear-Chaotic cliff approaching?)\nClear: deploy SOP; monitor. Complicated: experts; pick defensible alternative.\nComplex: parallel safe-to-fail probes; amplify wins. Chaotic: decisive action; re-diagnose.\nBoundary watch: shift signals | who monitors | re-diagnosis schedule\n```\n\n**Output template:**\n```\nCynefin Diagnosis: \nDomain: [Clear/Complicated/Complex/Chaotic/Confused] | Evidence: [cause-effect, expert agreement]\nMethod: [S-C-R / S-A-R / P-S-R / A-S-R] | Actions: | Mismatch cost (if any):\nBoundary watch: [shift signals | monitoring owner | re-diagnosis schedule]\n```\n\n*→ Method in Action: [Snowden at IBM (1999) and the HBR Synthesis (2007)](examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md) · [Apollo 13 Mission Response (1970)](examples/apollo-13-1970-mission-response.md)*\n*→ 2026 lens: [Sorting AI Decisions by Domain (2024–2026)](examples/sorting-ai-decisions-by-domain-2024-2026.md)*\n\n## Pack: Cynefin Domain Patterns\n\n| Domain | Examples | Method | Mistake |\n|---|---|---|---|\n| Clear | Routine compliance; manufacturing QC | S→Categorize→R; SOP | Over-analysis |\n| Complicated | Engineering design; surgery; M&A | S→Analyze→R; experts | Analysis paralysis |\n| Complex | Startup PMF; org culture; new market entry | Probe→S→R; safe-to-fail | Over-planning |\n| Chaotic | Crisis first 24h; security breach | Act→S→R; decisive action | Deliberating |\n| Confused | New market; leadership transition | Decompose; classify each | Defaulting to home domain |\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We just need a better plan\" | Often the issue is Complex — no plan works; probes and adaptation required. |\n| [D] \"Get me an expert\" | Right for Complicated. Wrong for Complex: experts disagree because cause-effect is emergent. |\n| [D] \"Do what worked last time\" | Right for Clear. Dangerous near Clear-Chaotic boundary — produces the cliff fall. |\n| [D] \"We need more data\" | Often a deflection in Complex/Chaotic where data only emerges from probes/action. |\n| [D] \"The plan is right; execution is the problem\" | Classic post-mortem rationalization when Complicated-domain plan failed on Complex-domain problem. |\n| [D] \"Best practices from industry X\" | Only transfers if industry X has the same domain structure. Complex ≠ Complicated. |\n| [D] \"We need more analysis / more decisiveness\" | Analysis: right for Complicated, wrong for Complex/Chaotic. Decisiveness: right for Chaotic, wrong for Complex. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Repeated failure always blamed on \"execution\" — \"Best practices\" imported without domain check\n- Experts disagree on the right answer (Complex signal) — Crisis response dominated by analysis\n- Complex situation managed with a single plan, not a probe portfolio\n- Team waiting for clarity in a domain where clarity only comes from acting\n\n## Verification\n\n- [ ] Domain explicitly named with diagnostic evidence\n- [ ] Decision method matched to domain (S-C-R / S-A-R / P-S-R / A-S-R)\n- [ ] If current approach mismatches: mismatch cost named\n- [ ] Boundary signals identified; re-diagnosis schedule set\n- [ ] If Complex: ≥3 safe-to-fail probes designed\n- [ ] If Chaotic: order-establishing action specified\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 227 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/cynefin** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\n*Agents: latest version & machine-readable metadata → https://www.deciqai.com/s/cynefin.json*\n\nFile v1.0.5:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"cynefin\",\n \"version\": \"1.0.5\",\n \"publishedAt\": 1784224600866\n}\n\nFile v1.0.5:references/sources.md\n\n# Sources — cynefin\n\n> *Primary sources for the [cynefin](../SKILL.md) skill.*\n\n- Snowden, D. J., & Boone, M. E. (2007). \"A leader's framework for decision making.\" *Harvard Business Review*, 85(11), 68-76. The mainstream introduction.\n- Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. Technical foundation.\n- Snowden, D. J. (2002). \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111.\n- Snowden, D. J., & Greenberg, R. (2020). *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge.\n- French, S. (2013). \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561.\n- Polanyi, M. (1966). *The Tacit Dimension.* Doubleday. (Background on tacit knowledge.)\n- Stacey, R. D. (1996). *Complexity and Creativity in Organizations.* Berrett-Koehler. (Complexity-science background.)\n- U.S. Army (2012). *Mission Command: Command and Control of Army Forces.* ADP 6-0. (Operational adoption.)\n- Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. (Apollo 13 example.)\n- NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA. (Apollo 13 example.)\n- Anthropic (2024–2025). \"Building Effective Agents\" and related public agent-engineering guidance. anthropic.com. (2024–2026 AI-decisions example — the \"start simple, add autonomy incrementally, keep humans in the loop, constrain the action space\" posture as Probe–Sense–Respond.)\n- Snowden, D. J., et al. (2020–2021). *Cynefin & Weaving Sense-Making into the Fabric of Our World* / Cynefin.io field guides. Cognitive Edge / The Cynefin Co. (Contemporary restatement of the domains and boundary dynamics used in the 2024–2026 AI example.)\n\nFile v1.0.5:examples/apollo-13-1970-mission-response.md\n\n# Method in Action: The Apollo 13 Mission Response (1970)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nApollo 13 is a compressed, four-day demonstration of every Cynefin domain — and of why survival depended on re-diagnosing the domain at each phase instead of running one method throughout.\n\n**The situation.** On April 13, 1970, roughly 56 hours into a routine lunar mission about 200,000 miles from Earth, oxygen tank 2 in the service module exploded, crippling the spacecraft's power, oxygen, and propulsion. Jim Lovell's report to Houston — \"Houston, we've had a problem\" — marked the instant the mission left its planned domain.\n\n**Phase 1 — Chaotic: Act–Sense–Respond.** In the first hour, cause-effect was absent: telemetry contradicted itself, controllers initially suspected instrumentation failure, oxygen was venting, and fuel cells were dying. There was no playbook for a dying service module en route to the Moon. Flight director Gene Kranz's move was the Chaotic-domain move: act first to establish order, not analyze. Mission Control powered down the command module Odyssey to preserve its re-entry batteries and moved the crew into the lunar module Aquarius as a lifeboat — an improvised, order-establishing action taken before anyone understood the cause. Deliberating would have cost the crew their margin. Only after order was established did the situation become sensible enough to re-classify.\n\n**Phase 2 — Complex: Probe–Sense–Respond.** With the crew stabilized, NASA faced novel problems with no procedures and emergent cause-effect: a lunar module designed to keep two men alive for two days now had to keep three alive for four; carbon dioxide was accumulating because the command module's square lithium hydroxide canisters did not fit the lunar module's round scrubber sockets; the power budget had no precedent. Experts could not simply compute the answer — ground teams probed. Engineers in Houston assembled a canister adapter from only the materials known to be aboard (plastic bags, cardboard, suit hose, tape), tested it, then read the build procedure up to the crew. Simulator teams ran candidate power-down and navigation configurations, sensed what held, amplified what worked, and discarded what failed. This was a portfolio of safe-to-fail experiments run on the ground so that failure would not be fatal in flight.\n\n**Phase 3 — Complicated: Sense–Analyze–Respond.** Some sub-problems were knowable with expertise and moved by analysis: the decision to use a free-return trajectory around the Moon rather than a risky direct abort, the PC+2 engine burn after lunar flyby to speed the return, and the command module power-up sequence — written and verified in simulators by astronaut Ken Mattingly and the ground team before being uplinked. Here the correct move was expert analysis of a knowable system, and NASA used it — but only for the sub-problems that genuinely lived in that domain.\n\n**Phase 4 — back to Clear.** After splashdown on April 17, 1970, NASA convened the Apollo 13 Review Board under Edgar Cortright. The board's Complicated-domain analysis traced the explosion to damaged wire insulation inside the oxygen tank, and its findings were converted into Clear-domain material: a redesigned tank, revised test procedures, and standard operating rules for later missions. The crisis's lessons were deliberately migrated down the domains until they became routine.\n\n**Mismatch cost, made visible.** The one moment the wrong-domain reflex surfaced was the abort debate: an immediate direct abort — turn the ship around now using the main engine — was the \"do what the contingency plan says\" answer. Analysis showed the main engine sat next to the damaged service module and could not be trusted; firing it was an unverifiable bet. Choosing the slower free-return trajectory was a domain-honest choice: it traded speed for a path whose cause-effect structure was actually knowable. Had Mission Control categorized the crisis as a standard abort scenario and fired the engine, the mismatch cost would likely have been the crew.\n\n**The misclassification watch.** The rescue worked because NASA never defaulted to its home domain. A nominal Apollo mission is Clear/Complicated territory — checklists and expert analysis — and the reflex to keep running the nominal playbook after the explosion would have been the classic Cynefin error. Instead, Mission Control treated domain diagnosis as continuous: chaotic action first, probes once stabilized, expert analysis for the knowable pieces, and codification afterward.\n\nThe mapped steps:\n\n1. **Describe:** Crippled spacecraft 200,000 miles out; current approach (nominal mission procedures) invalidated; stakeholders: three crew, Mission Control, contractor engineers.\n2. **Diagnose the domain:** First hour — cause-effect absent, telemetry incoherent → Chaotic. After stabilization — survival problems novel and emergent, experts unable to converge on answers without trials → Complex. Trajectory and re-entry — knowable with expertise → Complicated.\n3. **Match approach to domain:** Chaotic → Act–Sense–Respond (power down Odyssey, move crew to Aquarius). Complex → Probe–Sense–Respond (scrubber adapter and power procedures built and tested on the ground, wins amplified). Complicated → Sense–Analyze–Respond (free-return decision, PC+2 burn, power-up checklist from simulation).\n4. **Check boundary movement:** Continuous re-diagnosis as the situation shifted Chaotic → Complex → Complicated; post-mission Review Board pushed the lessons into the Clear domain as redesigns and SOPs.\n\nThe operational lesson: domains are phases, not labels. A single crisis can traverse the entire framework in days, and the team that survives is the one that keeps asking \"which domain are we in now?\" — not the one with the best plan for the domain it started in.\n\nPrimary sources: Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA.\n\nFile v1.0.5:examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md\n\n# Method in Action: Snowden at IBM (1999) and the HBR Synthesis (2007)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nDave Snowden developed the Cynefin framework while leading the Cynefin Centre for Organisational Complexity at IBM's Institute for Knowledge Management (1999-2004). The framework emerged from Snowden's work on knowledge management in complex organizations and his observation that classical management techniques — strategy frameworks, planning methodologies, KPI systems — kept failing when applied to certain kinds of problems. Snowden's diagnosis: those techniques were designed for Complicated problems (knowable cause-effect, expert analysis) and were being misapplied to Complex problems (emergent cause-effect, requires probing).\n\nThe framework had three formative influences:\n\n1. **Knowledge management at IBM.** Snowden observed that knowledge in complex organizations is largely tacit (Polanyi 1966) and contextually bound — it cannot be extracted and codified the way explicit knowledge can. This led to the recognition that knowledge transfer in Complex domains requires different methods than in Complicated.\n\n2. **Complexity science.** The Santa Fe Institute's work on complex adaptive systems (1980s-1990s) gave Snowden the mathematical foundation for the Complex domain: systems with many interacting agents, emergent properties, non-linear dynamics, and irreducible uncertainty.\n\n3. **Practical experience in conflict zones.** Snowden's later application of the framework in the British government's military planning (post-9/11), Singapore's risk-assessment programs, and complex civil-society interventions tested and refined the framework against high-stakes problems.\n\nThe 2007 *Harvard Business Review* paper, co-authored with Mary Boone, was the framework's mainstream introduction. It explicitly placed Cynefin in the management-decision context and provided concrete examples of each domain. The paper's most-cited passages:\n\n> \"Many executives are surprised when previously successful leadership approaches fail in new situations, but different contexts call for different kinds of responses. Before addressing a situation, leaders need to recognize which context governs it — and tailor their actions accordingly. Cynefin, which is Welsh for 'habitat,' encourages leaders to see things from new viewpoints, assimilate complex concepts, and address real-world problems and opportunities. It sorts the issues facing leaders into five contexts defined by the nature of the relationship between cause and effect. Four of these — simple, complicated, complex, and chaotic — require leaders to diagnose situations and act in contextually appropriate ways. The fifth — disorder — applies when it is unclear which of the other four contexts is predominant.\"\n>\n> — Snowden & Boone (2007), p. 69.\n\nThe paper also articulated the framework's most operationally important warning, the **danger of best-practice imposition in the wrong domain**:\n\n> \"The complex domain is where most of the high-stakes strategic decisions of contemporary organizations live. It is also where the most damaging managerial error occurs: the application of complicated-domain methods — exhaustive analysis, expert authority, comprehensive planning — to problems whose underlying dynamics are emergent. The result is a plan that looks impressive in PowerPoint, fails in execution, and generates a post-mortem that blames the execution rather than the underlying mismatch between method and problem.\"\n>\n> — Snowden & Boone (2007), pp. 72-73.\n\nThe paper's most cited example was the response to **Hurricane Katrina (2005)**: the initial response treated the situation as Complicated (apply the standard FEMA playbook), when it was actually Chaotic (the system had collapsed in ways that the playbook did not address). The delay in switching to chaotic-domain method (decisive action to establish order, then sense-respond) cost lives. Subsequent analysis of emergency-response failures in disasters (the COVID-19 pandemic early phase being a recent example) repeatedly shows the same pattern: organizations stuck in Complicated-domain methods when the situation has shifted to Complex or Chaotic.\n\nThe framework has been refined since 2007. Key updates:\n\n- **2014:** \"Simple\" renamed to \"Clear\" to avoid pejorative connotation.\n- **2019:** \"Aporetic\" introduced as a sub-region of Disorder — situations where you don't yet know but the not-knowing is itself the answer.\n- **2020:** \"Liminal\" domains introduced — transitional zones between the main domains, where the situation is moving from one to another.\n\nThe follow-up literature is large:\n\n**Kurtz, C. F., & Snowden, D. J. (2003).** \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. The technical foundation paper, predating the HBR synthesis.\n\n**Snowden, D. J. (2002).** \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111. The knowledge-management foundation.\n\n**Snowden, D. J., & Greenberg, R. (2020).** *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge. The comprehensive 2020 retrospective with updated framework and case studies.\n\n**Mowles, C. (2015).** *Managing in Uncertainty: Complexity and the Paradoxes of Everyday Organizational Life.* Routledge. Academic application to organizational management.\n\n**French, S. (2013).** \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561. The decision-analysis application.\n\nThe framework has shaped operational practice in multiple disciplines:\n\n**Military / defense.** The British and U.S. militaries have integrated Cynefin into officer-training and doctrine development. Counterinsurgency operations (Iraq 2005-2010, Afghanistan throughout) explicitly use the Complex-domain methodology of safe-to-fail experiments at scale. U.S. Army's Mission Command doctrine (2012-) is largely a Cynefin-influenced framework.\n\n**Software development.** Agile methodologies (Scrum, Kanban, Lean) are explicitly the Complex-domain response (Probe-Sense-Respond) to product development. Waterfall is the Complicated-domain method. The literature on \"agile transformation\" failures is largely the story of organizations attempting to import Complicated-domain governance onto Complex-domain work.\n\n**Healthcare.** Treating common illness (Clear), specialized diagnosis (Complicated), chronic-disease management with social determinants (Complex), and pandemic response (Chaotic transitioning to Complex) all require different methods. Healthcare quality movements that ignore the domain differences have produced poor results.\n\n**Public policy.** Wicked problems (climate change, drug policy, inequality, infrastructure) are paradigmatic Complex-domain issues. Treating them with Complicated-domain methods (expert commission, comprehensive plan, single national strategy) is the most-documented policy failure pattern in modern government. Snowden's framework has informed several national-government policy units (UK, Singapore, Australia).\n\n**Organizational change.** Culture change and strategic transformation are emergent, not predictable. Cynefin-influenced change methodologies (Liminal Action, SenseMaker) explicitly use safe-to-fail probes and multi-experiment portfolios instead of single-track transformation plans.\n\n**Crisis management.** First 24 hours of a crisis: Chaotic (Act-Sense-Respond). Next phase: Complex (Probe-Sense-Respond as stabilization continues). Later: Complicated (analyze what happened, build into doctrine). Modern crisis-management protocols increasingly recognize the domain-shift pattern.\n\nThree operational lessons from Cynefin:\n\n**First, the most expensive management error is using a Complicated-domain method on a Complex problem.** Symptoms: detailed plans that fail in execution, post-mortems that blame \"execution\" rather than method, recurrent failure of expert recommendations, \"best practices\" that don't work. The fix: switch to Probe-Sense-Respond, run multiple safe-to-fail experiments, amplify what works.\n\n**Second, the boundary between Clear and Chaotic is the most dangerous.** Long success in a Clear domain (routine operations, mature market, established practice) breeds complacency. A small change (new competitor, regulatory shift, technology disruption) can push the situation into Chaotic without warning. Leaders who respond with \"do what we always do\" preside over the cliff fall. The defense: explicit monitoring for boundary signals; sensors deliberately positioned to catch the transition.\n\n**Third, the framework is diagnostic, not prescriptive.** Cynefin does not tell you what to do; it tells you what *kind* of doing to do. The actual content of each domain's action (which experiments, which experts, which SOP, which decisive crisis intervention) requires domain expertise. Treating Cynefin as a recipe book reduces its value; treating it as a diagnostic lens preserves it.\n\nFile v1.0.5:examples/sorting-ai-decisions-by-domain-2024-2026.md\n\n# Method in Action: Sorting 2024–2026 AI Decisions by Domain\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nBy 2024–2026, many organizations building with AI had moved past \"should we use AI?\" into a harder question: *which* AI decisions can be run like engineering, which must be run like exploration, and which are outright fires. Cynefin's core warning — that the costliest error is treating a Complex problem as a Complicated one — maps almost exactly onto the mistakes teams made when they applied the same delivery playbook to a known API integration, to an autonomous agent, and to a live incident. This example walks one organization's AI portfolio through the skill's Process.\n\n**Step 1 — Describe.**\n\n- **Decision/situation:** A product org wants to ship three AI capabilities: (a) integrate a well-documented large language model API for text summarization, (b) build an \"agentic\" workflow where an LLM plans and calls tools autonomously across multi-step tasks, and (c) respond when that agent misbehaves in production.\n- **Current approach:** One delivery process for all three — scope it, estimate it, assign experts, ship on a roadmap date.\n- **What worked / not:** The integration shipped cleanly. The agentic workflow blew past estimates and behaved differently every time it was tested. Nobody had a plan for the production incident.\n- **Stakeholders:** Engineering, product, an ML/platform team, security/trust-and-safety, affected end users, executives allocating AI capex.\n\n**Step 2 — Diagnose the domain.** Run each capability through the diagnostic questions (5 experts converge? standard best practice works? interventions predictable?).\n\n- **(a) The known model integration — Complicated.** Cause-effect is knowable with expertise. Calling a documented API, handling rate limits and token budgets, adding retries and evals, caching, and streaming responses are all engineering problems with defensible right answers. Five competent engineers would broadly converge on the design. Best practice transfers. Interventions are predictable. This is not Clear (it takes genuine expertise, not an SOP), but it is knowable.\n- **(b) Agentic reliability and emergent behavior — Complex.** Here cause-effect is only visible in retrospect. An autonomous agent that plans, uses tools, and chains model calls exhibits behavior that emerges from the interaction of the model, the prompt, the tools, and the live environment — small changes to a prompt or a model version can shift outcomes unpredictably, and the same input can produce different trajectories. Experts genuinely *disagree* on the right architecture because the terrain is new and shifting. Standard software estimation fails. This is the classic Complex signal — and treating it as Complicated (just scope it, assign an expert, commit a date) is the exact mistake Cynefin flags as most costly.\n- **(c) A production incident — Chaotic.** When the deployed agent starts taking a harmful or runaway action against real users — leaking data, looping expensive tool calls, or emitting unsafe output at scale — cause-effect is in flux and every second of deliberation compounds harm. There is no time to analyze first.\n\n**Step 3 — Match approach to domain.**\n\n- **(a) Complicated → Sense–Analyze–Respond.** Put experienced engineers on it, evaluate a few defensible design alternatives (sync vs. streaming, which eval harness, prompt vs. fine-tune), pick one, and ship on a plan. Analysis is the right move; the only failure mode here is *analysis paralysis* — over-deliberating a solved-shape problem.\n- **(b) Complex → Probe–Sense–Respond.** Do **not** commit to one grand agent design on a roadmap date. Instead run a portfolio of **safe-to-fail probes**: constrain the agent's tool permissions, run it in a sandbox or shadow mode against real traffic without acting, add human-in-the-loop checkpoints, cap spend and action counts, and measure against evals. Sense which probes hold, **amplify** what works, and kill what fails. The output is not a prediction; it is a set of experiments cheap enough to be wrong. Publicly available 2024–2026 agent-building guidance from major AI labs converged on exactly this posture — start narrow, add autonomy incrementally, keep humans in the loop, and constrain the action space — which is Probe–Sense–Respond by another name.\n- **(c) Chaotic → Act–Sense–Respond.** Establish order first: trip the kill switch, revoke the agent's tool credentials, roll back to the previous model or disable the feature, and cut off the blast radius. Only once the bleeding stops do you re-classify — the post-incident investigation into *why* is then a Complicated (root-cause analysis) or Complex (emergent-behavior) problem, not a Chaotic one.\n\n**Step 4 — Check boundary movement + choose intervention.**\n\n- **Complicated → Complex under disruption.** The dangerous boundary is that a capability that *looks* Complicated can be Complex underneath. A model-version upgrade, a new tool added to the agent, or a shift in user behavior can move a \"solved\" integration back into emergent territory. The boundary watch: version-pinning and eval regressions as **shift signals**, the ML/platform team as the **monitoring owner**, and re-diagnosis on every model or tool change.\n- **The Clear–Chaotic cliff.** A seductive error in this period is treating a maturing AI feature as *Clear* — \"it worked in the demo, automate it and walk away.\" Complacent over-automation of an emergent system is precisely the fall off the Clear–Chaotic cliff: it looks routine right up until it produces a production incident. Keeping the feature classified as Complex (with probes, monitoring, and human oversight) rather than prematurely Clear is what prevents the cliff.\n- **Mismatch cost, named.** Running the agentic workflow (Complex) through the integration's process (Complicated) is what produced the blown estimates and the unhandled incident — the plan was \"right\" for a domain the problem did not live in. The observable symptom was the recurring \"the plan was fine, execution was the problem\" post-mortem, which Cynefin identifies as the tell of a Complicated-domain plan failing on a Complex-domain problem.\n\n**Output template, filled.**\n\n```\nCynefin Diagnosis: Enterprise AI portfolio, 2024–2026\nDomain: Complicated (known model integration) | Complex (agentic reliability) | Chaotic (production incident)\nEvidence: (a) experts converge, best practice transfers, interventions predictable\n (b) experts disagree, behavior emergent/retrospective, estimation fails\n (c) cause-effect in flux, harm compounds with delay\nMethod: (a) S-A-R (b) P-S-R (c) A-S-R\nActions: (a) expert design + evals, ship on plan\n (b) sandbox/shadow, constrained permissions, human-in-loop, capped spend, amplify winning probes\n (c) kill switch, revoke credentials, roll back, contain, then re-diagnose\nMismatch cost: running (b) as (a) → blown timelines + unhandled incident (the \"execution\" post-mortem)\nBoundary watch: model/tool version changes = shift signals | ML-platform team monitors | re-diagnose each release\n```\n\n**The lesson.** The AI decisions of 2024–2026 were not one problem to be managed with one methodology; they spanned three Cynefin domains that each demand a *different* decision method. Teams that treated agent-building as ordinary feature engineering paid the classic Complex-as-Complicated tax, while teams that ran constrained, safe-to-fail probes — and kept a kill switch ready for the Chaotic case — matched approach to domain. Domains are phases, not labels: as models and tools change, yesterday's Complicated integration can slide back into Complex, so the discipline is to keep asking \"which domain is this decision in *now?*\"\n\n*Sources: Snowden, D. J., & Boone, M. E. (2007). \"A Leader's Framework for Decision Making.\" Harvard Business Review, 85(11), 68–76 (Cynefin domains and the Complex-vs-Complicated error). Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy.\" IBM Systems Journal, 42(3), 462–483. Anthropic (2024–2025), \"Building effective agents\" and related public agent-engineering guidance (start simple, add autonomy incrementally, keep humans in the loop, constrain the action space). General characterizations of large-language-model and agent behavior — non-determinism, sensitivity to prompt/model changes, and emergent multi-step behavior — reflect widely documented, publicly reported properties as of early 2026.*\n\nFile v1.0.5:skill-card.md\n\n## Description:\n\nHelps agents diagnose decision contexts with the Cynefin framework and match actions to Clear, Complicated, Complex, Chaotic, or Confused domains.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[deciqai](https://clawhub.ai/user/deciqai)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nEmployees, external users, and developers use this skill to classify ambiguous decisions, crises, expert disagreements, and failed best-practice playbooks by Cynefin domain, then choose the matching decision method.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Users may over-rely on the skill's recommendations for irreversible business, operational, or crisis decisions.\n\nMitigation: Treat outputs as decision-support guidance and apply normal human review before taking high-impact action.\n\nRisk: Incorrect domain classification can lead to mismatched methods, such as over-planning a complex situation or over-analyzing a chaotic one.\n\nMitigation: Require the output to name diagnostic evidence, mismatch cost, boundary signals, and a re-diagnosis schedule before acting.\n\n## Reference(s):\n\n- [Cynefin skill page](https://clawhub.ai/deciqai/skills/cynefin)\n- [Primary sources](references/sources.md)\n- [Snowden at IBM and the HBR synthesis](examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md)\n- [Apollo 13 mission response](examples/apollo-13-1970-mission-response.md)\n- [Sorting AI decisions by domain](examples/sorting-ai-decisions-by-domain-2024-2026.md)\n- [deciqAI Cynefin page](https://www.deciqai.com/c/cynefin)\n- [deciqAI Cynefin metadata](https://www.deciqai.com/s/cynefin.json)\n\n## Skill Output:\n\n**Output Type(s):** [Text, Markdown, Guidance]\n\n**Output Format:** [Markdown]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May pause for user input during coach-mode steps and should be reviewed before irreversible business, operational, or crisis actions.]\n\n## Skill Version(s):\n\n1.0.5 (source: server release evidence)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nArchive v1.0.4: 7 files, 17758 bytes\n\nFiles: examples/apollo-13-1970-mission-response.md (6142b), examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md (9098b), examples/sorting-ai-decisions-by-domain-2024-2026.md (8583b), references/sources.md (1921b), skill-card.md (2667b), SKILL.md (7419b), _meta.json (126b)\n\nFile v1.0.4:SKILL.md\n\n---\nname: cynefin\ndescription: \"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know what to do first', 'best practice doesn't apply here', 'we need a different approach'. Do NOT activate when: situation is unambiguously routine (execution only); a specialized tool (OODA, expected value) already fits.\"\n---\n\n# Cynefin\n\n## Overview\n\n**Cynefin** (pronounced \"kuh-NEV-in\"; Welsh for \"habitat\") is a sense-making framework by Dave Snowden (IBM, 1999). Its claim: the right decision approach depends on which of five domains the situation falls into — **Clear** (obvious cause-effect, use SOP), **Complicated** (knowable with expertise, use analysis), **Complex** (emergent, probe first), **Chaotic** (absent cause-effect, act first), **Confused** (unknown domain, decompose first). The most common and costly error: treating Complex problems as Complicated.\n\nComposes with `ooda-loop`, `feedback-loops`, `antifragile`, `first-principles`.\n\n## When to Use\n\n- A familiar approach has stopped working and you can't articulate why\n- Experts disagree on the right answer — a crisis unfolding where the previous playbook doesn't apply\n- \"Best practices from X\" imported without checking if the domain matches\n- A team is over-planning something emergent, or \"let's get more data\" when data won't come without action\n- Allocating AI capex or racing AI-native competition: deciding which AI bets are engineering (Complicated), emergent agent/adoption experiments (Complex), or live incidents (Chaotic)\n\n**Not when:** domain is unambiguously Clear (execution only); small-stakes one-shot; specialized framework already fits.\n\n## Coaching Novices (Adaptive Front Door)\n\n**Engine mode:** concrete case → run The Process. **Coach mode:** unfamiliar → guide step by step.\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. Classify the situation (Clear/Complicated/Complex/Chaotic) and match decision method to domain.\n2. Check fit: if unambiguously routine (Clear), skip framework.\n3. Elicit their real case — decision, current method, cause-effect structure.\n> **[WAIT — do not advance until user responds]**\n4. Are cause-effect relationships obvious, knowable, emergent, or absent? Is current method matched?\n> **[WAIT — do not advance until user responds]**\n5. Close: named domain + matched decision method + boundary watch.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe:** `Decision/situation: | Current approach: | What worked/not: | Stakeholders:`\n\n### Step 2: Diagnose the domain\n```\nObvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)\n```\nDiagnostics: 5 experts converge? (Yes → Complicated; No → Complex). Standard best practice works? (Yes → Clear/Complicated; No → Complex/Chaotic). Interventions predictable? (Yes → Clear/Complicated; No → Complex/Chaotic).\n\n### Step 3: Match approach to domain\n```\nClear: Sense→Categorize→Respond (SOP/automate) | Complicated: Sense→Analyze→Respond (experts)\nComplex: Probe→Sense→Respond (safe-to-fail experiments, amplify wins)\nChaotic: Act→Sense→Respond (establish order, then re-classify) | Confused: decompose, classify each part\n```\n\n### Step 4: Check boundary movement + choose intervention\n```\nDomain shifted? (Complicated→Complex from disruption? Clear-Chaotic cliff approaching?)\nClear: deploy SOP; monitor. Complicated: experts; pick defensible alternative.\nComplex: parallel safe-to-fail probes; amplify wins. Chaotic: decisive action; re-diagnose.\nBoundary watch: shift signals | who monitors | re-diagnosis schedule\n```\n\n**Output template:**\n```\nCynefin Diagnosis: \nDomain: [Clear/Complicated/Complex/Chaotic/Confused] | Evidence: [cause-effect, expert agreement]\nMethod: [S-C-R / S-A-R / P-S-R / A-S-R] | Actions: | Mismatch cost (if any):\nBoundary watch: [shift signals | monitoring owner | re-diagnosis schedule]\n```\n\n*→ Method in Action: [Snowden at IBM (1999) and the HBR Synthesis (2007)](examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md) · [Apollo 13 Mission Response (1970)](examples/apollo-13-1970-mission-response.md)*\n*→ 2026 lens: [Sorting AI Decisions by Domain (2024–2026)](examples/sorting-ai-decisions-by-domain-2024-2026.md)*\n\n## Pack: Cynefin Domain Patterns\n\n| Domain | Examples | Method | Mistake |\n|---|---|---|---|\n| Clear | Routine compliance; manufacturing QC | S→Categorize→R; SOP | Over-analysis |\n| Complicated | Engineering design; surgery; M&A | S→Analyze→R; experts | Analysis paralysis |\n| Complex | Startup PMF; org culture; new market entry | Probe→S→R; safe-to-fail | Over-planning |\n| Chaotic | Crisis first 24h; security breach | Act→S→R; decisive action | Deliberating |\n| Confused | New market; leadership transition | Decompose; classify each | Defaulting to home domain |\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We just need a better plan\" | Often the issue is Complex — no plan works; probes and adaptation required. |\n| [D] \"Get me an expert\" | Right for Complicated. Wrong for Complex: experts disagree because cause-effect is emergent. |\n| [D] \"Do what worked last time\" | Right for Clear. Dangerous near Clear-Chaotic boundary — produces the cliff fall. |\n| [D] \"We need more data\" | Often a deflection in Complex/Chaotic where data only emerges from probes/action. |\n| [D] \"The plan is right; execution is the problem\" | Classic post-mortem rationalization when Complicated-domain plan failed on Complex-domain problem. |\n| [D] \"Best practices from industry X\" | Only transfers if industry X has the same domain structure. Complex ≠ Complicated. |\n| [D] \"We need more analysis / more decisiveness\" | Analysis: right for Complicated, wrong for Complex/Chaotic. Decisiveness: right for Chaotic, wrong for Complex. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Repeated failure always blamed on \"execution\" — \"Best practices\" imported without domain check\n- Experts disagree on the right answer (Complex signal) — Crisis response dominated by analysis\n- Complex situation managed with a single plan, not a probe portfolio\n- Team waiting for clarity in a domain where clarity only comes from acting\n\n## Verification\n\n- [ ] Domain explicitly named with diagnostic evidence\n- [ ] Decision method matched to domain (S-C-R / S-A-R / P-S-R / A-S-R)\n- [ ] If current approach mismatches: mismatch cost named\n- [ ] Boundary signals identified; re-diagnosis schedule set\n- [ ] If Complex: ≥3 safe-to-fail probes designed\n- [ ] If Chaotic: order-establishing action specified\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 189 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/cynefin** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.4:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"cynefin\",\n \"version\": \"1.0.4\",\n \"publishedAt\": 1783595805631\n}\n\nFile v1.0.4:references/sources.md\n\n# Sources — cynefin\n\n> *Primary sources for the [cynefin](../SKILL.md) skill.*\n\n- Snowden, D. J., & Boone, M. E. (2007). \"A leader's framework for decision making.\" *Harvard Business Review*, 85(11), 68-76. The mainstream introduction.\n- Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. Technical foundation.\n- Snowden, D. J. (2002). \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111.\n- Snowden, D. J., & Greenberg, R. (2020). *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge.\n- French, S. (2013). \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561.\n- Polanyi, M. (1966). *The Tacit Dimension.* Doubleday. (Background on tacit knowledge.)\n- Stacey, R. D. (1996). *Complexity and Creativity in Organizations.* Berrett-Koehler. (Complexity-science background.)\n- U.S. Army (2012). *Mission Command: Command and Control of Army Forces.* ADP 6-0. (Operational adoption.)\n- Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. (Apollo 13 example.)\n- NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA. (Apollo 13 example.)\n- Anthropic (2024–2025). \"Building Effective Agents\" and related public agent-engineering guidance. anthropic.com. (2024–2026 AI-decisions example — the \"start simple, add autonomy incrementally, keep humans in the loop, constrain the action space\" posture as Probe–Sense–Respond.)\n- Snowden, D. J., et al. (2020–2021). *Cynefin & Weaving Sense-Making into the Fabric of Our World* / Cynefin.io field guides. Cognitive Edge / The Cynefin Co. (Contemporary restatement of the domains and boundary dynamics used in the 2024–2026 AI example.)\n\nFile v1.0.4:examples/apollo-13-1970-mission-response.md\n\n# Method in Action: The Apollo 13 Mission Response (1970)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nApollo 13 is a compressed, four-day demonstration of every Cynefin domain — and of why survival depended on re-diagnosing the domain at each phase instead of running one method throughout.\n\n**The situation.** On April 13, 1970, roughly 56 hours into a routine lunar mission about 200,000 miles from Earth, oxygen tank 2 in the service module exploded, crippling the spacecraft's power, oxygen, and propulsion. Jim Lovell's report to Houston — \"Houston, we've had a problem\" — marked the instant the mission left its planned domain.\n\n**Phase 1 — Chaotic: Act–Sense–Respond.** In the first hour, cause-effect was absent: telemetry contradicted itself, controllers initially suspected instrumentation failure, oxygen was venting, and fuel cells were dying. There was no playbook for a dying service module en route to the Moon. Flight director Gene Kranz's move was the Chaotic-domain move: act first to establish order, not analyze. Mission Control powered down the command module Odyssey to preserve its re-entry batteries and moved the crew into the lunar module Aquarius as a lifeboat — an improvised, order-establishing action taken before anyone understood the cause. Deliberating would have cost the crew their margin. Only after order was established did the situation become sensible enough to re-classify.\n\n**Phase 2 — Complex: Probe–Sense–Respond.** With the crew stabilized, NASA faced novel problems with no procedures and emergent cause-effect: a lunar module designed to keep two men alive for two days now had to keep three alive for four; carbon dioxide was accumulating because the command module's square lithium hydroxide canisters did not fit the lunar module's round scrubber sockets; the power budget had no precedent. Experts could not simply compute the answer — ground teams probed. Engineers in Houston assembled a canister adapter from only the materials known to be aboard (plastic bags, cardboard, suit hose, tape), tested it, then read the build procedure up to the crew. Simulator teams ran candidate power-down and navigation configurations, sensed what held, amplified what worked, and discarded what failed. This was a portfolio of safe-to-fail experiments run on the ground so that failure would not be fatal in flight.\n\n**Phase 3 — Complicated: Sense–Analyze–Respond.** Some sub-problems were knowable with expertise and moved by analysis: the decision to use a free-return trajectory around the Moon rather than a risky direct abort, the PC+2 engine burn after lunar flyby to speed the return, and the command module power-up sequence — written and verified in simulators by astronaut Ken Mattingly and the ground team before being uplinked. Here the correct move was expert analysis of a knowable system, and NASA used it — but only for the sub-problems that genuinely lived in that domain.\n\n**Phase 4 — back to Clear.** After splashdown on April 17, 1970, NASA convened the Apollo 13 Review Board under Edgar Cortright. The board's Complicated-domain analysis traced the explosion to damaged wire insulation inside the oxygen tank, and its findings were converted into Clear-domain material: a redesigned tank, revised test procedures, and standard operating rules for later missions. The crisis's lessons were deliberately migrated down the domains until they became routine.\n\n**Mismatch cost, made visible.** The one moment the wrong-domain reflex surfaced was the abort debate: an immediate direct abort — turn the ship around now using the main engine — was the \"do what the contingency plan says\" answer. Analysis showed the main engine sat next to the damaged service module and could not be trusted; firing it was an unverifiable bet. Choosing the slower free-return trajectory was a domain-honest choice: it traded speed for a path whose cause-effect structure was actually knowable. Had Mission Control categorized the crisis as a standard abort scenario and fired the engine, the mismatch cost would likely have been the crew.\n\n**The misclassification watch.** The rescue worked because NASA never defaulted to its home domain. A nominal Apollo mission is Clear/Complicated territory — checklists and expert analysis — and the reflex to keep running the nominal playbook after the explosion would have been the classic Cynefin error. Instead, Mission Control treated domain diagnosis as continuous: chaotic action first, probes once stabilized, expert analysis for the knowable pieces, and codification afterward.\n\nThe mapped steps:\n\n1. **Describe:** Crippled spacecraft 200,000 miles out; current approach (nominal mission procedures) invalidated; stakeholders: three crew, Mission Control, contractor engineers.\n2. **Diagnose the domain:** First hour — cause-effect absent, telemetry incoherent → Chaotic. After stabilization — survival problems novel and emergent, experts unable to converge on answers without trials → Complex. Trajectory and re-entry — knowable with expertise → Complicated.\n3. **Match approach to domain:** Chaotic → Act–Sense–Respond (power down Odyssey, move crew to Aquarius). Complex → Probe–Sense–Respond (scrubber adapter and power procedures built and tested on the ground, wins amplified). Complicated → Sense–Analyze–Respond (free-return decision, PC+2 burn, power-up checklist from simulation).\n4. **Check boundary movement:** Continuous re-diagnosis as the situation shifted Chaotic → Complex → Complicated; post-mission Review Board pushed the lessons into the Clear domain as redesigns and SOPs.\n\nThe operational lesson: domains are phases, not labels. A single crisis can traverse the entire framework in days, and the team that survives is the one that keeps asking \"which domain are we in now?\" — not the one with the best plan for the domain it started in.\n\nPrimary sources: Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA.\n\nFile v1.0.4:examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md\n\n# Method in Action: Snowden at IBM (1999) and the HBR Synthesis (2007)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nDave Snowden developed the Cynefin framework while leading the Cynefin Centre for Organisational Complexity at IBM's Institute for Knowledge Management (1999-2004). The framework emerged from Snowden's work on knowledge management in complex organizations and his observation that classical management techniques — strategy frameworks, planning methodologies, KPI systems — kept failing when applied to certain kinds of problems. Snowden's diagnosis: those techniques were designed for Complicated problems (knowable cause-effect, expert analysis) and were being misapplied to Complex problems (emergent cause-effect, requires probing).\n\nThe framework had three formative influences:\n\n1. **Knowledge management at IBM.** Snowden observed that knowledge in complex organizations is largely tacit (Polanyi 1966) and contextually bound — it cannot be extracted and codified the way explicit knowledge can. This led to the recognition that knowledge transfer in Complex domains requires different methods than in Complicated.\n\n2. **Complexity science.** The Santa Fe Institute's work on complex adaptive systems (1980s-1990s) gave Snowden the mathematical foundation for the Complex domain: systems with many interacting agents, emergent properties, non-linear dynamics, and irreducible uncertainty.\n\n3. **Practical experience in conflict zones.** Snowden's later application of the framework in the British government's military planning (post-9/11), Singapore's risk-assessment programs, and complex civil-society interventions tested and refined the framework against high-stakes problems.\n\nThe 2007 *Harvard Business Review* paper, co-authored with Mary Boone, was the framework's mainstream introduction. It explicitly placed Cynefin in the management-decision context and provided concrete examples of each domain. The paper's most-cited passages:\n\n> \"Many executives are surprised when previously successful leadership approaches fail in new situations, but different contexts call for different kinds of responses. Before addressing a situation, leaders need to recognize which context governs it — and tailor their actions accordingly. Cynefin, which is Welsh for 'habitat,' encourages leaders to see things from new viewpoints, assimilate complex concepts, and address real-world problems and opportunities. It sorts the issues facing leaders into five contexts defined by the nature of the relationship between cause and effect. Four of these — simple, complicated, complex, and chaotic — require leaders to diagnose situations and act in contextually appropriate ways. The fifth — disorder — applies when it is unclear which of the other four contexts is predominant.\"\n>\n> — Snowden & Boone (2007), p. 69.\n\nThe paper also articulated the framework's most operationally important warning, the **danger of best-practice imposition in the wrong domain**:\n\n> \"The complex domain is where most of the high-stakes strategic decisions of contemporary organizations live. It is also where the most damaging managerial error occurs: the application of complicated-domain methods — exhaustive analysis, expert authority, comprehensive planning — to problems whose underlying dynamics are emergent. The result is a plan that looks impressive in PowerPoint, fails in execution, and generates a post-mortem that blames the execution rather than the underlying mismatch between method and problem.\"\n>\n> — Snowden & Boone (2007), pp. 72-73.\n\nThe paper's most cited example was the response to **Hurricane Katrina (2005)**: the initial response treated the situation as Complicated (apply the standard FEMA playbook), when it was actually Chaotic (the system had collapsed in ways that the playbook did not address). The delay in switching to chaotic-domain method (decisive action to establish order, then sense-respond) cost lives. Subsequent analysis of emergency-response failures in disasters (the COVID-19 pandemic early phase being a recent example) repeatedly shows the same pattern: organizations stuck in Complicated-domain methods when the situation has shifted to Complex or Chaotic.\n\nThe framework has been refined since 2007. Key updates:\n\n- **2014:** \"Simple\" renamed to \"Clear\" to avoid pejorative connotation.\n- **2019:** \"Aporetic\" introduced as a sub-region of Disorder — situations where you don't yet know but the not-knowing is itself the answer.\n- **2020:** \"Liminal\" domains introduced — transitional zones between the main domains, where the situation is moving from one to another.\n\nThe follow-up literature is large:\n\n**Kurtz, C. F., & Snowden, D. J. (2003).** \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. The technical foundation paper, predating the HBR synthesis.\n\n**Snowden, D. J. (2002).** \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111. The knowledge-management foundation.\n\n**Snowden, D. J., & Greenberg, R. (2020).** *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge. The comprehensive 2020 retrospective with updated framework and case studies.\n\n**Mowles, C. (2015).** *Managing in Uncertainty: Complexity and the Paradoxes of Everyday Organizational Life.* Routledge. Academic application to organizational management.\n\n**French, S. (2013).** \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561. The decision-analysis application.\n\nThe framework has shaped operational practice in multiple disciplines:\n\n**Military / defense.** The British and U.S. militaries have integrated Cynefin into officer-training and doctrine development. Counterinsurgency operations (Iraq 2005-2010, Afghanistan throughout) explicitly use the Complex-domain methodology of safe-to-fail experiments at scale. U.S. Army's Mission Command doctrine (2012-) is largely a Cynefin-influenced framework.\n\n**Software development.** Agile methodologies (Scrum, Kanban, Lean) are explicitly the Complex-domain response (Probe-Sense-Respond) to product development. Waterfall is the Complicated-domain method. The literature on \"agile transformation\" failures is largely the story of organizations attempting to import Complicated-domain governance onto Complex-domain work.\n\n**Healthcare.** Treating common illness (Clear), specialized diagnosis (Complicated), chronic-disease management with social determinants (Complex), and pandemic response (Chaotic transitioning to Complex) all require different methods. Healthcare quality movements that ignore the domain differences have produced poor results.\n\n**Public policy.** Wicked problems (climate change, drug policy, inequality, infrastructure) are paradigmatic Complex-domain issues. Treating them with Complicated-domain methods (expert commission, comprehensive plan, single national strategy) is the most-documented policy failure pattern in modern government. Snowden's framework has informed several national-government policy units (UK, Singapore, Australia).\n\n**Organizational change.** Culture change and strategic transformation are emergent, not predictable. Cynefin-influenced change methodologies (Liminal Action, SenseMaker) explicitly use safe-to-fail probes and multi-experiment portfolios instead of single-track transformation plans.\n\n**Crisis management.** First 24 hours of a crisis: Chaotic (Act-Sense-Respond). Next phase: Complex (Probe-Sense-Respond as stabilization continues). Later: Complicated (analyze what happened, build into doctrine). Modern crisis-management protocols increasingly recognize the domain-shift pattern.\n\nThree operational lessons from Cynefin:\n\n**First, the most expensive management error is using a Complicated-domain method on a Complex problem.** Symptoms: detailed plans that fail in execution, post-mortems that blame \"execution\" rather than method, recurrent failure of expert recommendations, \"best practices\" that don't work. The fix: switch to Probe-Sense-Respond, run multiple safe-to-fail experiments, amplify what works.\n\n**Second, the boundary between Clear and Chaotic is the most dangerous.** Long success in a Clear domain (routine operations, mature market, established practice) breeds complacency. A small change (new competitor, regulatory shift, technology disruption) can push the situation into Chaotic without warning. Leaders who respond with \"do what we always do\" preside over the cliff fall. The defense: explicit monitoring for boundary signals; sensors deliberately positioned to catch the transition.\n\n**Third, the framework is diagnostic, not prescriptive.** Cynefin does not tell you what to do; it tells you what *kind* of doing to do. The actual content of each domain's action (which experiments, which experts, which SOP, which decisive crisis intervention) requires domain expertise. Treating Cynefin as a recipe book reduces its value; treating it as a diagnostic lens preserves it.\n\nFile v1.0.4:examples/sorting-ai-decisions-by-domain-2024-2026.md\n\n# Method in Action: Sorting 2024–2026 AI Decisions by Domain\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nBy 2024–2026, many organizations building with AI had moved past \"should we use AI?\" into a harder question: *which* AI decisions can be run like engineering, which must be run like exploration, and which are outright fires. Cynefin's core warning — that the costliest error is treating a Complex problem as a Complicated one — maps almost exactly onto the mistakes teams made when they applied the same delivery playbook to a known API integration, to an autonomous agent, and to a live incident. This example walks one organization's AI portfolio through the skill's Process.\n\n**Step 1 — Describe.**\n\n- **Decision/situation:** A product org wants to ship three AI capabilities: (a) integrate a well-documented large language model API for text summarization, (b) build an \"agentic\" workflow where an LLM plans and calls tools autonomously across multi-step tasks, and (c) respond when that agent misbehaves in production.\n- **Current approach:** One delivery process for all three — scope it, estimate it, assign experts, ship on a roadmap date.\n- **What worked / not:** The integration shipped cleanly. The agentic workflow blew past estimates and behaved differently every time it was tested. Nobody had a plan for the production incident.\n- **Stakeholders:** Engineering, product, an ML/platform team, security/trust-and-safety, affected end users, executives allocating AI capex.\n\n**Step 2 — Diagnose the domain.** Run each capability through the diagnostic questions (5 experts converge? standard best practice works? interventions predictable?).\n\n- **(a) The known model integration — Complicated.** Cause-effect is knowable with expertise. Calling a documented API, handling rate limits and token budgets, adding retries and evals, caching, and streaming responses are all engineering problems with defensible right answers. Five competent engineers would broadly converge on the design. Best practice transfers. Interventions are predictable. This is not Clear (it takes genuine expertise, not an SOP), but it is knowable.\n- **(b) Agentic reliability and emergent behavior — Complex.** Here cause-effect is only visible in retrospect. An autonomous agent that plans, uses tools, and chains model calls exhibits behavior that emerges from the interaction of the model, the prompt, the tools, and the live environment — small changes to a prompt or a model version can shift outcomes unpredictably, and the same input can produce different trajectories. Experts genuinely *disagree* on the right architecture because the terrain is new and shifting. Standard software estimation fails. This is the classic Complex signal — and treating it as Complicated (just scope it, assign an expert, commit a date) is the exact mistake Cynefin flags as most costly.\n- **(c) A production incident — Chaotic.** When the deployed agent starts taking a harmful or runaway action against real users — leaking data, looping expensive tool calls, or emitting unsafe output at scale — cause-effect is in flux and every second of deliberation compounds harm. There is no time to analyze first.\n\n**Step 3 — Match approach to domain.**\n\n- **(a) Complicated → Sense–Analyze–Respond.** Put experienced engineers on it, evaluate a few defensible design alternatives (sync vs. streaming, which eval harness, prompt vs. fine-tune), pick one, and ship on a plan. Analysis is the right move; the only failure mode here is *analysis paralysis* — over-deliberating a solved-shape problem.\n- **(b) Complex → Probe–Sense–Respond.** Do **not** commit to one grand agent design on a roadmap date. Instead run a portfolio of **safe-to-fail probes**: constrain the agent's tool permissions, run it in a sandbox or shadow mode against real traffic without acting, add human-in-the-loop checkpoints, cap spend and action counts, and measure against evals. Sense which probes hold, **amplify** what works, and kill what fails. The output is not a prediction; it is a set of experiments cheap enough to be wrong. Publicly available 2024–2026 agent-building guidance from major AI labs converged on exactly this posture — start narrow, add autonomy incrementally, keep humans in the loop, and constrain the action space — which is Probe–Sense–Respond by another name.\n- **(c) Chaotic → Act–Sense–Respond.** Establish order first: trip the kill switch, revoke the agent's tool credentials, roll back to the previous model or disable the feature, and cut off the blast radius. Only once the bleeding stops do you re-classify — the post-incident investigation into *why* is then a Complicated (root-cause analysis) or Complex (emergent-behavior) problem, not a Chaotic one.\n\n**Step 4 — Check boundary movement + choose intervention.**\n\n- **Complicated → Complex under disruption.** The dangerous boundary is that a capability that *looks* Complicated can be Complex underneath. A model-version upgrade, a new tool added to the agent, or a shift in user behavior can move a \"solved\" integration back into emergent territory. The boundary watch: version-pinning and eval regressions as **shift signals**, the ML/platform team as the **monitoring owner**, and re-diagnosis on every model or tool change.\n- **The Clear–Chaotic cliff.** A seductive error in this period is treating a maturing AI feature as *Clear* — \"it worked in the demo, automate it and walk away.\" Complacent over-automation of an emergent system is precisely the fall off the Clear–Chaotic cliff: it looks routine right up until it produces a production incident. Keeping the feature classified as Complex (with probes, monitoring, and human oversight) rather than prematurely Clear is what prevents the cliff.\n- **Mismatch cost, named.** Running the agentic workflow (Complex) through the integration's process (Complicated) is what produced the blown estimates and the unhandled incident — the plan was \"right\" for a domain the problem did not live in. The observable symptom was the recurring \"the plan was fine, execution was the problem\" post-mortem, which Cynefin identifies as the tell of a Complicated-domain plan failing on a Complex-domain problem.\n\n**Output template, filled.**\n\n```\nCynefin Diagnosis: Enterprise AI portfolio, 2024–2026\nDomain: Complicated (known model integration) | Complex (agentic reliability) | Chaotic (production incident)\nEvidence: (a) experts converge, best practice transfers, interventions predictable\n (b) experts disagree, behavior emergent/retrospective, estimation fails\n (c) cause-effect in flux, harm compounds with delay\nMethod: (a) S-A-R (b) P-S-R (c) A-S-R\nActions: (a) expert design + evals, ship on plan\n (b) sandbox/shadow, constrained permissions, human-in-loop, capped spend, amplify winning probes\n (c) kill switch, revoke credentials, roll back, contain, then re-diagnose\nMismatch cost: running (b) as (a) → blown timelines + unhandled incident (the \"execution\" post-mortem)\nBoundary watch: model/tool version changes = shift signals | ML-platform team monitors | re-diagnose each release\n```\n\n**The lesson.** The AI decisions of 2024–2026 were not one problem to be managed with one methodology; they spanned three Cynefin domains that each demand a *different* decision method. Teams that treated agent-building as ordinary feature engineering paid the classic Complex-as-Complicated tax, while teams that ran constrained, safe-to-fail probes — and kept a kill switch ready for the Chaotic case — matched approach to domain. Domains are phases, not labels: as models and tools change, yesterday's Complicated integration can slide back into Complex, so the discipline is to keep asking \"which domain is this decision in *now?*\"\n\n*Sources: Snowden, D. J., & Boone, M. E. (2007). \"A Leader's Framework for Decision Making.\" Harvard Business Review, 85(11), 68–76 (Cynefin domains and the Complex-vs-Complicated error). Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy.\" IBM Systems Journal, 42(3), 462–483. Anthropic (2024–2025), \"Building effective agents\" and related public agent-engineering guidance (start simple, add autonomy incrementally, keep humans in the loop, constrain the action space). General characterizations of large-language-model and agent behavior — non-determinism, sensitivity to prompt/model changes, and emergent multi-step behavior — reflect widely documented, publicly reported properties as of early 2026.*\n\nFile v1.0.4:skill-card.md\n\n## Description:
\nActivate when best practices keep failing, experts disagree, an old playbook no longer works, a crisis lacks an obvious first move, or a decision needs a different approach; do not activate for unambiguously routine execution or when a more specialized framework already fits.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nEmployees, external users, developers, and decision makers use this skill to classify uncertain situations into Cynefin domains and choose the matching decision method. It is especially useful when plans, expert advice, or inherited best practices are failing because the problem is complex or chaotic rather than clear or complicated.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Users may over-rely on the skill for real crisis response, business change, credential revocation, spending decisions, or operational actions.
\nMitigation: Keep human judgment and explicit authorization in the loop before taking real-world action.
\nRisk: A Cynefin classification can be wrong or incomplete when the user provides an underspecified situation.
\nMitigation: Use the skill's diagnostic questions, boundary watch, and re-diagnosis schedule before committing to an intervention.
\n\n\n## Reference(s):
\n- [Cynefin Skill Page](https://clawhub.ai/deciqai/skills/cynefin)
\n- [Sources - cynefin](artifact/references/sources.md)
\n- [Snowden at IBM and the HBR Synthesis](artifact/examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md)
\n- [Apollo 13 Mission Response](artifact/examples/apollo-13-1970-mission-response.md)
\n- [Sorting AI Decisions by Domain](artifact/examples/sorting-ai-decisions-by-domain-2024-2026.md)
\n- [deciqAI Cynefin Runtime Page](https://www.deciqai.com/c/cynefin)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, guidance]
\n**Output Format:** [Markdown diagnosis and coaching prompts with structured decision-method recommendations]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [May stop at explicit WAIT checkpoints when coaching a novice through a real case.]
\n\n## Skill Version(s):
\n1.0.4 (source: release evidence)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v1.0.3: 6 files, 13125 bytes\n\nFiles: examples/apollo-13-1970-mission-response.md (6142b), examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md (9098b), references/sources.md (1367b), skill-card.md (2374b), SKILL.md (7117b), _meta.json (126b)\n\nFile v1.0.3:SKILL.md\n\n---\nname: cynefin\ndescription: \"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know what to do first', 'best practice doesn't apply here', 'we need a different approach'. Do NOT activate when: situation is unambiguously routine (execution only); a specialized tool (OODA, expected value) already fits.\"\n---\n\n# Cynefin\n\n## Overview\n\n**Cynefin** (pronounced \"kuh-NEV-in\"; Welsh for \"habitat\") is a sense-making framework by Dave Snowden (IBM, 1999). Its claim: the right decision approach depends on which of five domains the situation falls into — **Clear** (obvious cause-effect, use SOP), **Complicated** (knowable with expertise, use analysis), **Complex** (emergent, probe first), **Chaotic** (absent cause-effect, act first), **Confused** (unknown domain, decompose first). The most common and costly error: treating Complex problems as Complicated.\n\nComposes with `ooda-loop`, `feedback-loops`, `antifragile`, `first-principles`.\n\n## When to Use\n\n- A familiar approach has stopped working and you can't articulate why\n- Experts disagree on the right answer — a crisis unfolding where the previous playbook doesn't apply\n- \"Best practices from X\" imported without checking if the domain matches\n- A team is over-planning something emergent, or \"let's get more data\" when data won't come without action\n\n**Not when:** domain is unambiguously Clear (execution only); small-stakes one-shot; specialized framework already fits.\n\n## Coaching Novices (Adaptive Front Door)\n\n**Engine mode:** concrete case → run The Process. **Coach mode:** unfamiliar → guide step by step.\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. Classify the situation (Clear/Complicated/Complex/Chaotic) and match decision method to domain.\n2. Check fit: if unambiguously routine (Clear), skip framework.\n3. Elicit their real case — decision, current method, cause-effect structure.\n> **[WAIT — do not advance until user responds]**\n4. Are cause-effect relationships obvious, knowable, emergent, or absent? Is current method matched?\n> **[WAIT — do not advance until user responds]**\n5. Close: named domain + matched decision method + boundary watch.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe:** `Decision/situation: | Current approach: | What worked/not: | Stakeholders:`\n\n### Step 2: Diagnose the domain\n```\nObvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)\n```\nDiagnostics: 5 experts converge? (Yes → Complicated; No → Complex). Standard best practice works? (Yes → Clear/Complicated; No → Complex/Chaotic). Interventions predictable? (Yes → Clear/Complicated; No → Complex/Chaotic).\n\n### Step 3: Match approach to domain\n```\nClear: Sense→Categorize→Respond (SOP/automate) | Complicated: Sense→Analyze→Respond (experts)\nComplex: Probe→Sense→Respond (safe-to-fail experiments, amplify wins)\nChaotic: Act→Sense→Respond (establish order, then re-classify) | Confused: decompose, classify each part\n```\n\n### Step 4: Check boundary movement + choose intervention\n```\nDomain shifted? (Complicated→Complex from disruption? Clear-Chaotic cliff approaching?)\nClear: deploy SOP; monitor. Complicated: experts; pick defensible alternative.\nComplex: parallel safe-to-fail probes; amplify wins. Chaotic: decisive action; re-diagnose.\nBoundary watch: shift signals | who monitors | re-diagnosis schedule\n```\n\n**Output template:**\n```\nCynefin Diagnosis: \nDomain: [Clear/Complicated/Complex/Chaotic/Confused] | Evidence: [cause-effect, expert agreement]\nMethod: [S-C-R / S-A-R / P-S-R / A-S-R] | Actions: | Mismatch cost (if any):\nBoundary watch: [shift signals | monitoring owner | re-diagnosis schedule]\n```\n\n*→ Method in Action: [Snowden at IBM (1999) and the HBR Synthesis (2007)](examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md) · [Apollo 13 Mission Response (1970)](examples/apollo-13-1970-mission-response.md)*\n\n## Pack: Cynefin Domain Patterns\n\n| Domain | Examples | Method | Mistake |\n|---|---|---|---|\n| Clear | Routine compliance; manufacturing QC | S→Categorize→R; SOP | Over-analysis |\n| Complicated | Engineering design; surgery; M&A | S→Analyze→R; experts | Analysis paralysis |\n| Complex | Startup PMF; org culture; new market entry | Probe→S→R; safe-to-fail | Over-planning |\n| Chaotic | Crisis first 24h; security breach | Act→S→R; decisive action | Deliberating |\n| Confused | New market; leadership transition | Decompose; classify each | Defaulting to home domain |\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We just need a better plan\" | Often the issue is Complex — no plan works; probes and adaptation required. |\n| [D] \"Get me an expert\" | Right for Complicated. Wrong for Complex: experts disagree because cause-effect is emergent. |\n| [D] \"Do what worked last time\" | Right for Clear. Dangerous near Clear-Chaotic boundary — produces the cliff fall. |\n| [D] \"We need more data\" | Often a deflection in Complex/Chaotic where data only emerges from probes/action. |\n| [D] \"The plan is right; execution is the problem\" | Classic post-mortem rationalization when Complicated-domain plan failed on Complex-domain problem. |\n| [D] \"Best practices from industry X\" | Only transfers if industry X has the same domain structure. Complex ≠ Complicated. |\n| [D] \"We need more analysis / more decisiveness\" | Analysis: right for Complicated, wrong for Complex/Chaotic. Decisiveness: right for Chaotic, wrong for Complex. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Repeated failure always blamed on \"execution\" — \"Best practices\" imported without domain check\n- Experts disagree on the right answer (Complex signal) — Crisis response dominated by analysis\n- Complex situation managed with a single plan, not a probe portfolio\n- Team waiting for clarity in a domain where clarity only comes from acting\n\n## Verification\n\n- [ ] Domain explicitly named with diagnostic evidence\n- [ ] Decision method matched to domain (S-C-R / S-A-R / P-S-R / A-S-R)\n- [ ] If current approach mismatches: mismatch cost named\n- [ ] Boundary signals identified; re-diagnosis schedule set\n- [ ] If Complex: ≥3 safe-to-fail probes designed\n- [ ] If Chaotic: order-establishing action specified\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 164 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/cynefin** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.3:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"cynefin\",\n \"version\": \"1.0.3\",\n \"publishedAt\": 1783508315880\n}\n\nFile v1.0.3:references/sources.md\n\n# Sources — cynefin\n\n> *Primary sources for the [cynefin](../SKILL.md) skill.*\n\n- Snowden, D. J., & Boone, M. E. (2007). \"A leader's framework for decision making.\" *Harvard Business Review*, 85(11), 68-76. The mainstream introduction.\n- Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. Technical foundation.\n- Snowden, D. J. (2002). \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111.\n- Snowden, D. J., & Greenberg, R. (2020). *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge.\n- French, S. (2013). \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561.\n- Polanyi, M. (1966). *The Tacit Dimension.* Doubleday. (Background on tacit knowledge.)\n- Stacey, R. D. (1996). *Complexity and Creativity in Organizations.* Berrett-Koehler. (Complexity-science background.)\n- U.S. Army (2012). *Mission Command: Command and Control of Army Forces.* ADP 6-0. (Operational adoption.)\n- Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. (Apollo 13 example.)\n- NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA. (Apollo 13 example.)\n\nFile v1.0.3:examples/apollo-13-1970-mission-response.md\n\n# Method in Action: The Apollo 13 Mission Response (1970)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nApollo 13 is a compressed, four-day demonstration of every Cynefin domain — and of why survival depended on re-diagnosing the domain at each phase instead of running one method throughout.\n\n**The situation.** On April 13, 1970, roughly 56 hours into a routine lunar mission about 200,000 miles from Earth, oxygen tank 2 in the service module exploded, crippling the spacecraft's power, oxygen, and propulsion. Jim Lovell's report to Houston — \"Houston, we've had a problem\" — marked the instant the mission left its planned domain.\n\n**Phase 1 — Chaotic: Act–Sense–Respond.** In the first hour, cause-effect was absent: telemetry contradicted itself, controllers initially suspected instrumentation failure, oxygen was venting, and fuel cells were dying. There was no playbook for a dying service module en route to the Moon. Flight director Gene Kranz's move was the Chaotic-domain move: act first to establish order, not analyze. Mission Control powered down the command module Odyssey to preserve its re-entry batteries and moved the crew into the lunar module Aquarius as a lifeboat — an improvised, order-establishing action taken before anyone understood the cause. Deliberating would have cost the crew their margin. Only after order was established did the situation become sensible enough to re-classify.\n\n**Phase 2 — Complex: Probe–Sense–Respond.** With the crew stabilized, NASA faced novel problems with no procedures and emergent cause-effect: a lunar module designed to keep two men alive for two days now had to keep three alive for four; carbon dioxide was accumulating because the command module's square lithium hydroxide canisters did not fit the lunar module's round scrubber sockets; the power budget had no precedent. Experts could not simply compute the answer — ground teams probed. Engineers in Houston assembled a canister adapter from only the materials known to be aboard (plastic bags, cardboard, suit hose, tape), tested it, then read the build procedure up to the crew. Simulator teams ran candidate power-down and navigation configurations, sensed what held, amplified what worked, and discarded what failed. This was a portfolio of safe-to-fail experiments run on the ground so that failure would not be fatal in flight.\n\n**Phase 3 — Complicated: Sense–Analyze–Respond.** Some sub-problems were knowable with expertise and moved by analysis: the decision to use a free-return trajectory around the Moon rather than a risky direct abort, the PC+2 engine burn after lunar flyby to speed the return, and the command module power-up sequence — written and verified in simulators by astronaut Ken Mattingly and the ground team before being uplinked. Here the correct move was expert analysis of a knowable system, and NASA used it — but only for the sub-problems that genuinely lived in that domain.\n\n**Phase 4 — back to Clear.** After splashdown on April 17, 1970, NASA convened the Apollo 13 Review Board under Edgar Cortright. The board's Complicated-domain analysis traced the explosion to damaged wire insulation inside the oxygen tank, and its findings were converted into Clear-domain material: a redesigned tank, revised test procedures, and standard operating rules for later missions. The crisis's lessons were deliberately migrated down the domains until they became routine.\n\n**Mismatch cost, made visible.** The one moment the wrong-domain reflex surfaced was the abort debate: an immediate direct abort — turn the ship around now using the main engine — was the \"do what the contingency plan says\" answer. Analysis showed the main engine sat next to the damaged service module and could not be trusted; firing it was an unverifiable bet. Choosing the slower free-return trajectory was a domain-honest choice: it traded speed for a path whose cause-effect structure was actually knowable. Had Mission Control categorized the crisis as a standard abort scenario and fired the engine, the mismatch cost would likely have been the crew.\n\n**The misclassification watch.** The rescue worked because NASA never defaulted to its home domain. A nominal Apollo mission is Clear/Complicated territory — checklists and expert analysis — and the reflex to keep running the nominal playbook after the explosion would have been the classic Cynefin error. Instead, Mission Control treated domain diagnosis as continuous: chaotic action first, probes once stabilized, expert analysis for the knowable pieces, and codification afterward.\n\nThe mapped steps:\n\n1. **Describe:** Crippled spacecraft 200,000 miles out; current approach (nominal mission procedures) invalidated; stakeholders: three crew, Mission Control, contractor engineers.\n2. **Diagnose the domain:** First hour — cause-effect absent, telemetry incoherent → Chaotic. After stabilization — survival problems novel and emergent, experts unable to converge on answers without trials → Complex. Trajectory and re-entry — knowable with expertise → Complicated.\n3. **Match approach to domain:** Chaotic → Act–Sense–Respond (power down Odyssey, move crew to Aquarius). Complex → Probe–Sense–Respond (scrubber adapter and power procedures built and tested on the ground, wins amplified). Complicated → Sense–Analyze–Respond (free-return decision, PC+2 burn, power-up checklist from simulation).\n4. **Check boundary movement:** Continuous re-diagnosis as the situation shifted Chaotic → Complex → Complicated; post-mission Review Board pushed the lessons into the Clear domain as redesigns and SOPs.\n\nThe operational lesson: domains are phases, not labels. A single crisis can traverse the entire framework in days, and the team that survives is the one that keeps asking \"which domain are we in now?\" — not the one with the best plan for the domain it started in.\n\nPrimary sources: Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA.\n\nFile v1.0.3:examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md\n\n# Method in Action: Snowden at IBM (1999) and the HBR Synthesis (2007)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nDave Snowden developed the Cynefin framework while leading the Cynefin Centre for Organisational Complexity at IBM's Institute for Knowledge Management (1999-2004). The framework emerged from Snowden's work on knowledge management in complex organizations and his observation that classical management techniques — strategy frameworks, planning methodologies, KPI systems — kept failing when applied to certain kinds of problems. Snowden's diagnosis: those techniques were designed for Complicated problems (knowable cause-effect, expert analysis) and were being misapplied to Complex problems (emergent cause-effect, requires probing).\n\nThe framework had three formative influences:\n\n1. **Knowledge management at IBM.** Snowden observed that knowledge in complex organizations is largely tacit (Polanyi 1966) and contextually bound — it cannot be extracted and codified the way explicit knowledge can. This led to the recognition that knowledge transfer in Complex domains requires different methods than in Complicated.\n\n2. **Complexity science.** The Santa Fe Institute's work on complex adaptive systems (1980s-1990s) gave Snowden the mathematical foundation for the Complex domain: systems with many interacting agents, emergent properties, non-linear dynamics, and irreducible uncertainty.\n\n3. **Practical experience in conflict zones.** Snowden's later application of the framework in the British government's military planning (post-9/11), Singapore's risk-assessment programs, and complex civil-society interventions tested and refined the framework against high-stakes problems.\n\nThe 2007 *Harvard Business Review* paper, co-authored with Mary Boone, was the framework's mainstream introduction. It explicitly placed Cynefin in the management-decision context and provided concrete examples of each domain. The paper's most-cited passages:\n\n> \"Many executives are surprised when previously successful leadership approaches fail in new situations, but different contexts call for different kinds of responses. Before addressing a situation, leaders need to recognize which context governs it — and tailor their actions accordingly. Cynefin, which is Welsh for 'habitat,' encourages leaders to see things from new viewpoints, assimilate complex concepts, and address real-world problems and opportunities. It sorts the issues facing leaders into five contexts defined by the nature of the relationship between cause and effect. Four of these — simple, complicated, complex, and chaotic — require leaders to diagnose situations and act in contextually appropriate ways. The fifth — disorder — applies when it is unclear which of the other four contexts is predominant.\"\n>\n> — Snowden & Boone (2007), p. 69.\n\nThe paper also articulated the framework's most operationally important warning, the **danger of best-practice imposition in the wrong domain**:\n\n> \"The complex domain is where most of the high-stakes strategic decisions of contemporary organizations live. It is also where the most damaging managerial error occurs: the application of complicated-domain methods — exhaustive analysis, expert authority, comprehensive planning — to problems whose underlying dynamics are emergent. The result is a plan that looks impressive in PowerPoint, fails in execution, and generates a post-mortem that blames the execution rather than the underlying mismatch between method and problem.\"\n>\n> — Snowden & Boone (2007), pp. 72-73.\n\nThe paper's most cited example was the response to **Hurricane Katrina (2005)**: the initial response treated the situation as Complicated (apply the standard FEMA playbook), when it was actually Chaotic (the system had collapsed in ways that the playbook did not address). The delay in switching to chaotic-domain method (decisive action to establish order, then sense-respond) cost lives. Subsequent analysis of emergency-response failures in disasters (the COVID-19 pandemic early phase being a recent example) repeatedly shows the same pattern: organizations stuck in Complicated-domain methods when the situation has shifted to Complex or Chaotic.\n\nThe framework has been refined since 2007. Key updates:\n\n- **2014:** \"Simple\" renamed to \"Clear\" to avoid pejorative connotation.\n- **2019:** \"Aporetic\" introduced as a sub-region of Disorder — situations where you don't yet know but the not-knowing is itself the answer.\n- **2020:** \"Liminal\" domains introduced — transitional zones between the main domains, where the situation is moving from one to another.\n\nThe follow-up literature is large:\n\n**Kurtz, C. F., & Snowden, D. J. (2003).** \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. The technical foundation paper, predating the HBR synthesis.\n\n**Snowden, D. J. (2002).** \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111. The knowledge-management foundation.\n\n**Snowden, D. J., & Greenberg, R. (2020).** *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge. The comprehensive 2020 retrospective with updated framework and case studies.\n\n**Mowles, C. (2015).** *Managing in Uncertainty: Complexity and the Paradoxes of Everyday Organizational Life.* Routledge. Academic application to organizational management.\n\n**French, S. (2013).** \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561. The decision-analysis application.\n\nThe framework has shaped operational practice in multiple disciplines:\n\n**Military / defense.** The British and U.S. militaries have integrated Cynefin into officer-training and doctrine development. Counterinsurgency operations (Iraq 2005-2010, Afghanistan throughout) explicitly use the Complex-domain methodology of safe-to-fail experiments at scale. U.S. Army's Mission Command doctrine (2012-) is largely a Cynefin-influenced framework.\n\n**Software development.** Agile methodologies (Scrum, Kanban, Lean) are explicitly the Complex-domain response (Probe-Sense-Respond) to product development. Waterfall is the Complicated-domain method. The literature on \"agile transformation\" failures is largely the story of organizations attempting to import Complicated-domain governance onto Complex-domain work.\n\n**Healthcare.** Treating common illness (Clear), specialized diagnosis (Complicated), chronic-disease management with social determinants (Complex), and pandemic response (Chaotic transitioning to Complex) all require different methods. Healthcare quality movements that ignore the domain differences have produced poor results.\n\n**Public policy.** Wicked problems (climate change, drug policy, inequality, infrastructure) are paradigmatic Complex-domain issues. Treating them with Complicated-domain methods (expert commission, comprehensive plan, single national strategy) is the most-documented policy failure pattern in modern government. Snowden's framework has informed several national-government policy units (UK, Singapore, Australia).\n\n**Organizational change.** Culture change and strategic transformation are emergent, not predictable. Cynefin-influenced change methodologies (Liminal Action, SenseMaker) explicitly use safe-to-fail probes and multi-experiment portfolios instead of single-track transformation plans.\n\n**Crisis management.** First 24 hours of a crisis: Chaotic (Act-Sense-Respond). Next phase: Complex (Probe-Sense-Respond as stabilization continues). Later: Complicated (analyze what happened, build into doctrine). Modern crisis-management protocols increasingly recognize the domain-shift pattern.\n\nThree operational lessons from Cynefin:\n\n**First, the most expensive management error is using a Complicated-domain method on a Complex problem.** Symptoms: detailed plans that fail in execution, post-mortems that blame \"execution\" rather than method, recurrent failure of expert recommendations, \"best practices\" that don't work. The fix: switch to Probe-Sense-Respond, run multiple safe-to-fail experiments, amplify what works.\n\n**Second, the boundary between Clear and Chaotic is the most dangerous.** Long success in a Clear domain (routine operations, mature market, established practice) breeds complacency. A small change (new competitor, regulatory shift, technology disruption) can push the situation into Chaotic without warning. Leaders who respond with \"do what we always do\" preside over the cliff fall. The defense: explicit monitoring for boundary signals; sensors deliberately positioned to catch the transition.\n\n**Third, the framework is diagnostic, not prescriptive.** Cynefin does not tell you what to do; it tells you what *kind* of doing to do. The actual content of each domain's action (which experiments, which experts, which SOP, which decisive crisis intervention) requires domain expertise. Treating Cynefin as a recipe book reduces its value; treating it as a diagnostic lens preserves it.\n\nFile v1.0.3:skill-card.md\n\n## Description:
\nCynefin helps an agent diagnose whether a situation is Clear, Complicated, Complex, Chaotic, or Confused and match the decision method to that domain.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nEmployees, external users, and agents use this skill to classify ambiguous, failing, or crisis decisions by Cynefin domain and choose a matching response pattern. It is intended for sense-making when best practices, expert consensus, or existing playbooks no longer fit the situation.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: The skill may be used for high-stakes planning or crisis framing where an incorrect domain diagnosis could influence real-world actions.
\nMitigation: Treat outputs as advisory decision support, require human review, and involve appropriate domain experts before acting.
\nRisk: Users may apply the framework to routine decisions or cases where a specialized tool already fits better.
\nMitigation: Apply the skill's not-when checks and verify the domain evidence before using the recommended response pattern.
\n\n\n## Reference(s):
\n- [Sources - cynefin](artifact/references/sources.md)
\n- [Snowden at IBM (1999) and the HBR Synthesis (2007)](artifact/examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md)
\n- [Apollo 13 Mission Response (1970)](artifact/examples/apollo-13-1970-mission-response.md)
\n- [Cynefin Skill Page](https://clawhub.ai/deciqai/skills/cynefin)
\n\n\n## Skill Output:
\n**Output Type(s):** [guidance, markdown, text]
\n**Output Format:** [Markdown diagnosis with structured decision guidance]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [Includes a named domain, diagnostic evidence, matched decision method, actions, mismatch cost when relevant, and a boundary watch.]
\n\n## Skill Version(s):
\n1.0.3 (source: server release metadata)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v1.0.2: 6 files, 13061 bytes\n\nFiles: examples/apollo-13-1970-mission-response.md (6142b), examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md (9098b), references/sources.md (1367b), skill-card.md (2073b), SKILL.md (7214b), _meta.json (126b)\n\nFile v1.0.2:SKILL.md\n\n---\nname: cynefin\ndescription: \"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know what to do first', 'best practice doesn't apply here', 'we need a different approach'. Do NOT activate when: situation is unambiguously routine (execution only); a specialized tool (OODA, expected value) already fits.\"\n---\n\n# Cynefin\n\n## Overview\n\n**Cynefin** (pronounced \"kuh-NEV-in\"; Welsh for \"habitat\") is a sense-making framework by Dave Snowden (IBM, 1999). Its claim: the right decision approach depends on which of five domains the situation falls into — **Clear** (obvious cause-effect, use SOP), **Complicated** (knowable with expertise, use analysis), **Complex** (emergent, probe first), **Chaotic** (absent cause-effect, act first), **Confused** (unknown domain, decompose first). The most common and costly error: treating Complex problems as Complicated.\n\nComposes with `ooda-loop`, `feedback-loops`, `antifragile`, `first-principles`.\n\n## When to Use\n\n- A familiar approach has stopped working and you can't articulate why\n- Experts disagree on the right answer — a crisis unfolding where the previous playbook doesn't apply\n- \"Best practices from X\" imported without checking if the domain matches\n- A team is over-planning something emergent, or \"let's get more data\" when data won't come without action\n\n**Not when:** domain is unambiguously Clear (execution only); small-stakes one-shot; specialized framework already fits.\n\n## Coaching Novices (Adaptive Front Door)\n\n**Engine mode:** concrete case → run The Process. **Coach mode:** unfamiliar → guide step by step.\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. Classify the situation (Clear/Complicated/Complex/Chaotic) and match decision method to domain.\n2. Check fit: if unambiguously routine (Clear), skip framework.\n3. Elicit their real case — decision, current method, cause-effect structure.\n> **[WAIT — do not advance until user responds]**\n4. Are cause-effect relationships obvious, knowable, emergent, or absent? Is current method matched?\n> **[WAIT — do not advance until user responds]**\n5. Close: named domain + matched decision method + boundary watch.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe:** `Decision/situation: | Current approach: | What worked/not: | Stakeholders:`\n\n### Step 2: Diagnose the domain\n```\nObvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)\n```\nDiagnostics: 5 experts converge? (Yes → Complicated; No → Complex). Standard best practice works? (Yes → Clear/Complicated; No → Complex/Chaotic). Interventions predictable? (Yes → Clear/Complicated; No → Complex/Chaotic).\n\n### Step 3: Match approach to domain\n```\nClear: Sense→Categorize→Respond (SOP/automate) | Complicated: Sense→Analyze→Respond (experts)\nComplex: Probe→Sense→Respond (safe-to-fail experiments, amplify wins)\nChaotic: Act→Sense→Respond (establish order, then re-classify) | Confused: decompose, classify each part\n```\n\n### Step 4: Check boundary movement + choose intervention\n```\nDomain shifted? (Complicated→Complex from disruption? Clear-Chaotic cliff approaching?)\nClear: deploy SOP; monitor. Complicated: experts; pick defensible alternative.\nComplex: parallel safe-to-fail probes; amplify wins. Chaotic: decisive action; re-diagnose.\nBoundary watch: shift signals | who monitors | re-diagnosis schedule\n```\n\n**Output template:**\n```\nCynefin Diagnosis: \nDomain: [Clear/Complicated/Complex/Chaotic/Confused] | Evidence: [cause-effect, expert agreement]\nMethod: [S-C-R / S-A-R / P-S-R / A-S-R] | Actions: | Mismatch cost (if any):\nBoundary watch: [shift signals | monitoring owner | re-diagnosis schedule]\n```\n\n*→ Method in Action: [Snowden at IBM (1999) and the HBR Synthesis (2007)](examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md) · [Apollo 13 Mission Response (1970)](examples/apollo-13-1970-mission-response.md)*\n\n## Pack: Cynefin Domain Patterns\n\n| Domain | Examples | Method | Mistake |\n|---|---|---|---|\n| Clear | Routine compliance; manufacturing QC | S→Categorize→R; SOP | Over-analysis |\n| Complicated | Engineering design; surgery; M&A | S→Analyze→R; experts | Analysis paralysis |\n| Complex | Startup PMF; org culture; new market entry | Probe→S→R; safe-to-fail | Over-planning |\n| Chaotic | Crisis first 24h; security breach | Act→S→R; decisive action | Deliberating |\n| Confused | New market; leadership transition | Decompose; classify each | Defaulting to home domain |\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We just need a better plan\" | Often the issue is Complex — no plan works; probes and adaptation required. |\n| [D] \"Get me an expert\" | Right for Complicated. Wrong for Complex: experts disagree because cause-effect is emergent. |\n| [D] \"Do what worked last time\" | Right for Clear. Dangerous near Clear-Chaotic boundary — produces the cliff fall. |\n| [D] \"We need more data\" | Often a deflection in Complex/Chaotic where data only emerges from probes/action. |\n| [D] \"The plan is right; execution is the problem\" | Classic post-mortem rationalization when Complicated-domain plan failed on Complex-domain problem. |\n| [D] \"Best practices from industry X\" | Only transfers if industry X has the same domain structure. Complex ≠ Complicated. |\n| [D] \"We need more analysis / more decisiveness\" | Analysis: right for Complicated, wrong for Complex/Chaotic. Decisiveness: right for Chaotic, wrong for Complex. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Repeated failure always blamed on \"execution\" — \"Best practices\" imported without domain check\n- Experts disagree on the right answer (Complex signal) — Crisis response dominated by analysis\n- Complex situation managed with a single plan, not a probe portfolio\n- Team waiting for clarity in a domain where clarity only comes from acting\n\n## Verification\n\n- [ ] Domain explicitly named with diagnostic evidence\n- [ ] Decision method matched to domain (S-C-R / S-A-R / P-S-R / A-S-R)\n- [ ] If current approach mismatches: mismatch cost named\n- [ ] Boundary signals identified; re-diagnosis schedule set\n- [ ] If Complex: ≥3 safe-to-fail probes designed\n- [ ] If Chaotic: order-establishing action specified\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 163 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/skills/cynefin?utm_source=clawhub&utm_medium=marketplace&utm_campaign=knowledge-skills&utm_content=cynefin** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.2:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"cynefin\",\n \"version\": \"1.0.2\",\n \"publishedAt\": 1783471409472\n}\n\nFile v1.0.2:references/sources.md\n\n# Sources — cynefin\n\n> *Primary sources for the [cynefin](../SKILL.md) skill.*\n\n- Snowden, D. J., & Boone, M. E. (2007). \"A leader's framework for decision making.\" *Harvard Business Review*, 85(11), 68-76. The mainstream introduction.\n- Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. Technical foundation.\n- Snowden, D. J. (2002). \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111.\n- Snowden, D. J., & Greenberg, R. (2020). *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge.\n- French, S. (2013). \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561.\n- Polanyi, M. (1966). *The Tacit Dimension.* Doubleday. (Background on tacit knowledge.)\n- Stacey, R. D. (1996). *Complexity and Creativity in Organizations.* Berrett-Koehler. (Complexity-science background.)\n- U.S. Army (2012). *Mission Command: Command and Control of Army Forces.* ADP 6-0. (Operational adoption.)\n- Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. (Apollo 13 example.)\n- NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA. (Apollo 13 example.)\n\nFile v1.0.2:examples/apollo-13-1970-mission-response.md\n\n# Method in Action: The Apollo 13 Mission Response (1970)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nApollo 13 is a compressed, four-day demonstration of every Cynefin domain — and of why survival depended on re-diagnosing the domain at each phase instead of running one method throughout.\n\n**The situation.** On April 13, 1970, roughly 56 hours into a routine lunar mission about 200,000 miles from Earth, oxygen tank 2 in the service module exploded, crippling the spacecraft's power, oxygen, and propulsion. Jim Lovell's report to Houston — \"Houston, we've had a problem\" — marked the instant the mission left its planned domain.\n\n**Phase 1 — Chaotic: Act–Sense–Respond.** In the first hour, cause-effect was absent: telemetry contradicted itself, controllers initially suspected instrumentation failure, oxygen was venting, and fuel cells were dying. There was no playbook for a dying service module en route to the Moon. Flight director Gene Kranz's move was the Chaotic-domain move: act first to establish order, not analyze. Mission Control powered down the command module Odyssey to preserve its re-entry batteries and moved the crew into the lunar module Aquarius as a lifeboat — an improvised, order-establishing action taken before anyone understood the cause. Deliberating would have cost the crew their margin. Only after order was established did the situation become sensible enough to re-classify.\n\n**Phase 2 — Complex: Probe–Sense–Respond.** With the crew stabilized, NASA faced novel problems with no procedures and emergent cause-effect: a lunar module designed to keep two men alive for two days now had to keep three alive for four; carbon dioxide was accumulating because the command module's square lithium hydroxide canisters did not fit the lunar module's round scrubber sockets; the power budget had no precedent. Experts could not simply compute the answer — ground teams probed. Engineers in Houston assembled a canister adapter from only the materials known to be aboard (plastic bags, cardboard, suit hose, tape), tested it, then read the build procedure up to the crew. Simulator teams ran candidate power-down and navigation configurations, sensed what held, amplified what worked, and discarded what failed. This was a portfolio of safe-to-fail experiments run on the ground so that failure would not be fatal in flight.\n\n**Phase 3 — Complicated: Sense–Analyze–Respond.** Some sub-problems were knowable with expertise and moved by analysis: the decision to use a free-return trajectory around the Moon rather than a risky direct abort, the PC+2 engine burn after lunar flyby to speed the return, and the command module power-up sequence — written and verified in simulators by astronaut Ken Mattingly and the ground team before being uplinked. Here the correct move was expert analysis of a knowable system, and NASA used it — but only for the sub-problems that genuinely lived in that domain.\n\n**Phase 4 — back to Clear.** After splashdown on April 17, 1970, NASA convened the Apollo 13 Review Board under Edgar Cortright. The board's Complicated-domain analysis traced the explosion to damaged wire insulation inside the oxygen tank, and its findings were converted into Clear-domain material: a redesigned tank, revised test procedures, and standard operating rules for later missions. The crisis's lessons were deliberately migrated down the domains until they became routine.\n\n**Mismatch cost, made visible.** The one moment the wrong-domain reflex surfaced was the abort debate: an immediate direct abort — turn the ship around now using the main engine — was the \"do what the contingency plan says\" answer. Analysis showed the main engine sat next to the damaged service module and could not be trusted; firing it was an unverifiable bet. Choosing the slower free-return trajectory was a domain-honest choice: it traded speed for a path whose cause-effect structure was actually knowable. Had Mission Control categorized the crisis as a standard abort scenario and fired the engine, the mismatch cost would likely have been the crew.\n\n**The misclassification watch.** The rescue worked because NASA never defaulted to its home domain. A nominal Apollo mission is Clear/Complicated territory — checklists and expert analysis — and the reflex to keep running the nominal playbook after the explosion would have been the classic Cynefin error. Instead, Mission Control treated domain diagnosis as continuous: chaotic action first, probes once stabilized, expert analysis for the knowable pieces, and codification afterward.\n\nThe mapped steps:\n\n1. **Describe:** Crippled spacecraft 200,000 miles out; current approach (nominal mission procedures) invalidated; stakeholders: three crew, Mission Control, contractor engineers.\n2. **Diagnose the domain:** First hour — cause-effect absent, telemetry incoherent → Chaotic. After stabilization — survival problems novel and emergent, experts unable to converge on answers without trials → Complex. Trajectory and re-entry — knowable with expertise → Complicated.\n3. **Match approach to domain:** Chaotic → Act–Sense–Respond (power down Odyssey, move crew to Aquarius). Complex → Probe–Sense–Respond (scrubber adapter and power procedures built and tested on the ground, wins amplified). Complicated → Sense–Analyze–Respond (free-return decision, PC+2 burn, power-up checklist from simulation).\n4. **Check boundary movement:** Continuous re-diagnosis as the situation shifted Chaotic → Complex → Complicated; post-mission Review Board pushed the lessons into the Clear domain as redesigns and SOPs.\n\nThe operational lesson: domains are phases, not labels. A single crisis can traverse the entire framework in days, and the team that survives is the one that keeps asking \"which domain are we in now?\" — not the one with the best plan for the domain it started in.\n\nPrimary sources: Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA.\n\nFile v1.0.2:examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md\n\n# Method in Action: Snowden at IBM (1999) and the HBR Synthesis (2007)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nDave Snowden developed the Cynefin framework while leading the Cynefin Centre for Organisational Complexity at IBM's Institute for Knowledge Management (1999-2004). The framework emerged from Snowden's work on knowledge management in complex organizations and his observation that classical management techniques — strategy frameworks, planning methodologies, KPI systems — kept failing when applied to certain kinds of problems. Snowden's diagnosis: those techniques were designed for Complicated problems (knowable cause-effect, expert analysis) and were being misapplied to Complex problems (emergent cause-effect, requires probing).\n\nThe framework had three formative influences:\n\n1. **Knowledge management at IBM.** Snowden observed that knowledge in complex organizations is largely tacit (Polanyi 1966) and contextually bound — it cannot be extracted and codified the way explicit knowledge can. This led to the recognition that knowledge transfer in Complex domains requires different methods than in Complicated.\n\n2. **Complexity science.** The Santa Fe Institute's work on complex adaptive systems (1980s-1990s) gave Snowden the mathematical foundation for the Complex domain: systems with many interacting agents, emergent properties, non-linear dynamics, and irreducible uncertainty.\n\n3. **Practical experience in conflict zones.** Snowden's later application of the framework in the British government's military planning (post-9/11), Singapore's risk-assessment programs, and complex civil-society interventions tested and refined the framework against high-stakes problems.\n\nThe 2007 *Harvard Business Review* paper, co-authored with Mary Boone, was the framework's mainstream introduction. It explicitly placed Cynefin in the management-decision context and provided concrete examples of each domain. The paper's most-cited passages:\n\n> \"Many executives are surprised when previously successful leadership approaches fail in new situations, but different contexts call for different kinds of responses. Before addressing a situation, leaders need to recognize which context governs it — and tailor their actions accordingly. Cynefin, which is Welsh for 'habitat,' encourages leaders to see things from new viewpoints, assimilate complex concepts, and address real-world problems and opportunities. It sorts the issues facing leaders into five contexts defined by the nature of the relationship between cause and effect. Four of these — simple, complicated, complex, and chaotic — require leaders to diagnose situations and act in contextually appropriate ways. The fifth — disorder — applies when it is unclear which of the other four contexts is predominant.\"\n>\n> — Snowden & Boone (2007), p. 69.\n\nThe paper also articulated the framework's most operationally important warning, the **danger of best-practice imposition in the wrong domain**:\n\n> \"The complex domain is where most of the high-stakes strategic decisions of contemporary organizations live. It is also where the most damaging managerial error occurs: the application of complicated-domain methods — exhaustive analysis, expert authority, comprehensive planning — to problems whose underlying dynamics are emergent. The result is a plan that looks impressive in PowerPoint, fails in execution, and generates a post-mortem that blames the execution rather than the underlying mismatch between method and problem.\"\n>\n> — Snowden & Boone (2007), pp. 72-73.\n\nThe paper's most cited example was the response to **Hurricane Katrina (2005)**: the initial response treated the situation as Complicated (apply the standard FEMA playbook), when it was actually Chaotic (the system had collapsed in ways that the playbook did not address). The delay in switching to chaotic-domain method (decisive action to establish order, then sense-respond) cost lives. Subsequent analysis of emergency-response failures in disasters (the COVID-19 pandemic early phase being a recent example) repeatedly shows the same pattern: organizations stuck in Complicated-domain methods when the situation has shifted to Complex or Chaotic.\n\nThe framework has been refined since 2007. Key updates:\n\n- **2014:** \"Simple\" renamed to \"Clear\" to avoid pejorative connotation.\n- **2019:** \"Aporetic\" introduced as a sub-region of Disorder — situations where you don't yet know but the not-knowing is itself the answer.\n- **2020:** \"Liminal\" domains introduced — transitional zones between the main domains, where the situation is moving from one to another.\n\nThe follow-up literature is large:\n\n**Kurtz, C. F., & Snowden, D. J. (2003).** \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. The technical foundation paper, predating the HBR synthesis.\n\n**Snowden, D. J. (2002).** \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111. The knowledge-management foundation.\n\n**Snowden, D. J., & Greenberg, R. (2020).** *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge. The comprehensive 2020 retrospective with updated framework and case studies.\n\n**Mowles, C. (2015).** *Managing in Uncertainty: Complexity and the Paradoxes of Everyday Organizational Life.* Routledge. Academic application to organizational management.\n\n**French, S. (2013).** \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561. The decision-analysis application.\n\nThe framework has shaped operational practice in multiple disciplines:\n\n**Military / defense.** The British and U.S. militaries have integrated Cynefin into officer-training and doctrine development. Counterinsurgency operations (Iraq 2005-2010, Afghanistan throughout) explicitly use the Complex-domain methodology of safe-to-fail experiments at scale. U.S. Army's Mission Command doctrine (2012-) is largely a Cynefin-influenced framework.\n\n**Software development.** Agile methodologies (Scrum, Kanban, Lean) are explicitly the Complex-domain response (Probe-Sense-Respond) to product development. Waterfall is the Complicated-domain method. The literature on \"agile transformation\" failures is largely the story of organizations attempting to import Complicated-domain governance onto Complex-domain work.\n\n**Healthcare.** Treating common illness (Clear), specialized diagnosis (Complicated), chronic-disease management with social determinants (Complex), and pandemic response (Chaotic transitioning to Complex) all require different methods. Healthcare quality movements that ignore the domain differences have produced poor results.\n\n**Public policy.** Wicked problems (climate change, drug policy, inequality, infrastructure) are paradigmatic Complex-domain issues. Treating them with Complicated-domain methods (expert commission, comprehensive plan, single national strategy) is the most-documented policy failure pattern in modern government. Snowden's framework has informed several national-government policy units (UK, Singapore, Australia).\n\n**Organizational change.** Culture change and strategic transformation are emergent, not predictable. Cynefin-influenced change methodologies (Liminal Action, SenseMaker) explicitly use safe-to-fail probes and multi-experiment portfolios instead of single-track transformation plans.\n\n**Crisis management.** First 24 hours of a crisis: Chaotic (Act-Sense-Respond). Next phase: Complex (Probe-Sense-Respond as stabilization continues). Later: Complicated (analyze what happened, build into doctrine). Modern crisis-management protocols increasingly recognize the domain-shift pattern.\n\nThree operational lessons from Cynefin:\n\n**First, the most expensive management error is using a Complicated-domain method on a Complex problem.** Symptoms: detailed plans that fail in execution, post-mortems that blame \"execution\" rather than method, recurrent failure of expert recommendations, \"best practices\" that don't work. The fix: switch to Probe-Sense-Respond, run multiple safe-to-fail experiments, amplify what works.\n\n**Second, the boundary between Clear and Chaotic is the most dangerous.** Long success in a Clear domain (routine operations, mature market, established practice) breeds complacency. A small change (new competitor, regulatory shift, technology disruption) can push the situation into Chaotic without warning. Leaders who respond with \"do what we always do\" preside over the cliff fall. The defense: explicit monitoring for boundary signals; sensors deliberately positioned to catch the transition.\n\n**Third, the framework is diagnostic, not prescriptive.** Cynefin does not tell you what to do; it tells you what *kind* of doing to do. The actual content of each domain's action (which experiments, which experts, which SOP, which decisive crisis intervention) requires domain expertise. Treating Cynefin as a recipe book reduces its value; treating it as a diagnostic lens preserves it.\n\nFile v1.0.2:skill-card.md\n\n## Description:
\nCynefin helps agents diagnose whether a decision situation is clear, complicated, complex, chaotic, or confused and match the response method accordingly.
\n\nThis skill is ready for commercial/non-commercial use.
\n\n## Publisher:
\n[deciqai](https://clawhub.ai/user/deciqai)
\n\n### License/Terms of Use:
\nMIT-0
\n\n\n## Use Case:
\nDevelopers, strategists, and operators use this skill as decision-support guidance when established playbooks fail, experts disagree, or a crisis requires matching actions to the situation's Cynefin domain. It should support human judgment rather than authorize real crisis, business, legal, financial, safety, or irreversible actions.
\n\n### Deployment Geography for Use:
\nGlobal
\n\n## Known Risks and Mitigations:
\nRisk: Review before execution as proposals could introduce incorrect or misleading guidance into skills.
\nMitigation: Review and scan skill before deployment.
\n\n## Reference(s):
\n- [ClawHub Skill Page](https://clawhub.ai/deciqai/skills/cynefin)
\n- [Primary Sources](references/sources.md)
\n- [Snowden at IBM and HBR Synthesis Example](examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md)
\n- [Apollo 13 Mission Response Example](examples/apollo-13-1970-mission-response.md)
\n\n\n## Skill Output:
\n**Output Type(s):** [text, markdown, guidance]
\n**Output Format:** [Markdown diagnostic guidance with structured questions, domain classification, action recommendations, and a boundary-watch checklist]
\n**Output Parameters:** [1D]
\n**Other Properties Related to Output:** [Non-executable advisory output; no code, shell commands, or configuration are produced.]
\n\n## Skill Version(s):
\n1.0.2 (source: server evidence release.version)
\n\n## Ethical Considerations:
\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
\n\nArchive v1.0.1: 5 files, 9719 bytes\n\nFiles: examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md (9098b), references/sources.md (1129b), skill-card.md (2009b), SKILL.md (7002b), _meta.json (126b)\n\nFile v1.0.1:SKILL.md\n\n---\nname: cynefin\ndescription: \"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know what to do first', 'best practice doesn't apply here', 'we need a different approach'. Do NOT activate when: situation is unambiguously routine (execution only); a specialized tool (OODA, expected value) already fits.\"\n---\n\n# Cynefin\n\n## Overview\n\n**Cynefin** (pronounced \"kuh-NEV-in\"; Welsh for \"habitat\") is a sense-making framework by Dave Snowden (IBM, 1999). Its claim: the right decision approach depends on which of five domains the situation falls into — **Clear** (obvious cause-effect, use SOP), **Complicated** (knowable with expertise, use analysis), **Complex** (emergent, probe first), **Chaotic** (absent cause-effect, act first), **Confused** (unknown domain, decompose first). The most common and costly error: treating Complex problems as Complicated.\n\nComposes with [`ooda-loop`](../ooda-loop/SKILL.md), [`feedback-loops`](../feedback-loops/SKILL.md), [`antifragile`](../antifragile/SKILL.md), [`first-principles`](../first-principles/SKILL.md).\n\n## When to Use\n\n- A familiar approach has stopped working and you can't articulate why\n- Experts disagree on the right answer — a crisis unfolding where the previous playbook doesn't apply\n- \"Best practices from X\" imported without checking if the domain matches\n- A team is over-planning something emergent, or \"let's get more data\" when data won't come without action\n\n**Not when:** domain is unambiguously Clear (execution only); small-stakes one-shot; specialized framework already fits.\n\n## Coaching Novices (Adaptive Front Door)\n\n**Engine mode:** concrete case → run The Process. **Coach mode:** unfamiliar → guide step by step.\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. Classify the situation (Clear/Complicated/Complex/Chaotic) and match decision method to domain.\n2. Check fit: if unambiguously routine (Clear), skip framework.\n3. Elicit their real case — decision, current method, cause-effect structure.\n> **[WAIT — do not advance until user responds]**\n4. Are cause-effect relationships obvious, knowable, emergent, or absent? Is current method matched?\n> **[WAIT — do not advance until user responds]**\n5. Close: named domain + matched decision method + boundary watch.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe:** `Decision/situation: | Current approach: | What worked/not: | Stakeholders:`\n\n### Step 2: Diagnose the domain\n```\nObvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)\n```\nDiagnostics: 5 experts converge? (Yes → Complicated; No → Complex). Standard best practice works? (Yes → Clear/Complicated; No → Complex/Chaotic). Interventions predictable? (Yes → Clear/Complicated; No → Complex/Chaotic).\n\n### Step 3: Match approach to domain\n```\nClear: Sense→Categorize→Respond (SOP/automate) | Complicated: Sense→Analyze→Respond (experts)\nComplex: Probe→Sense→Respond (safe-to-fail experiments, amplify wins)\nChaotic: Act→Sense→Respond (establish order, then re-classify) | Confused: decompose, classify each part\n```\n\n### Step 4: Check boundary movement + choose intervention\n```\nDomain shifted? (Complicated→Complex from disruption? Clear-Chaotic cliff approaching?)\nClear: deploy SOP; monitor. Complicated: experts; pick defensible alternative.\nComplex: parallel safe-to-fail probes; amplify wins. Chaotic: decisive action; re-diagnose.\nBoundary watch: shift signals | who monitors | re-diagnosis schedule\n```\n\n**Output template:**\n```\nCynefin Diagnosis: \nDomain: [Clear/Complicated/Complex/Chaotic/Confused] | Evidence: [cause-effect, expert agreement]\nMethod: [S-C-R / S-A-R / P-S-R / A-S-R] | Actions: | Mismatch cost (if any):\nBoundary watch: [shift signals | monitoring owner | re-diagnosis schedule]\n```\n\n*→ Method in Action: [Snowden at IBM (1999) and the HBR Synthesis (2007)](examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md)*\n\n## Pack: Cynefin Domain Patterns\n\n| Domain | Examples | Method | Mistake |\n|---|---|---|---|\n| Clear | Routine compliance; manufacturing QC | S→Categorize→R; SOP | Over-analysis |\n| Complicated | Engineering design; surgery; M&A | S→Analyze→R; experts | Analysis paralysis |\n| Complex | Startup PMF; org culture; new market entry | Probe→S→R; safe-to-fail | Over-planning |\n| Chaotic | Crisis first 24h; security breach | Act→S→R; decisive action | Deliberating |\n| Confused | New market; leadership transition | Decompose; classify each | Defaulting to home domain |\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We just need a better plan\" | Often the issue is Complex — no plan works; probes and adaptation required. |\n| [D] \"Get me an expert\" | Right for Complicated. Wrong for Complex: experts disagree because cause-effect is emergent. |\n| [D] \"Do what worked last time\" | Right for Clear. Dangerous near Clear-Chaotic boundary — produces the cliff fall. |\n| [D] \"We need more data\" | Often a deflection in Complex/Chaotic where data only emerges from probes/action. |\n| [D] \"The plan is right; execution is the problem\" | Classic post-mortem rationalization when Complicated-domain plan failed on Complex-domain problem. |\n| [D] \"Best practices from industry X\" | Only transfers if industry X has the same domain structure. Complex ≠ Complicated. |\n| [D] \"We need more analysis / more decisiveness\" | Analysis: right for Complicated, wrong for Complex/Chaotic. Decisiveness: right for Chaotic, wrong for Complex. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Repeated failure always blamed on \"execution\" — \"Best practices\" imported without domain check\n- Experts disagree on the right answer (Complex signal) — Crisis response dominated by analysis\n- Complex situation managed with a single plan, not a probe portfolio\n- Team waiting for clarity in a domain where clarity only comes from acting\n\n## Verification\n\n- [ ] Domain explicitly named with diagnostic evidence\n- [ ] Decision method matched to domain (S-C-R / S-A-R / P-S-R / A-S-R)\n- [ ] If current approach mismatches: mismatch cost named\n- [ ] Boundary signals identified; re-diagnosis schedule set\n- [ ] If Complex: ≥3 safe-to-fail probes designed\n- [ ] If Chaotic: order-establishing action specified\n\n---\n\n*Part of **deciqAI Knowledge Skills** — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.*\n\nFile v1.0.1:_meta.json\n\n{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"cynefin\",\n \"version\": \"1.0.1\",\n \"publishedAt\": 1783456315489\n}\n\nFile v1.0.1:references/sources.md\n\n# Sources — cynefin\n\n> *Primary sources for the [cynefin](../SKILL.md) skill.*\n\n- Snowden, D. J., & Boone, M. E. (2007). \"A leader's framework for decision making.\" *Harvard Business Review*, 85(11), 68-76. The mainstream introduction.\n- Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. Technical foundation.\n- Snowden, D. J. (2002). \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111.\n- Snowden, D. J., & Greenberg, R. (2020). *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge.\n- French, S. (2013). \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561.\n- Polanyi, M. (1966). *The Tacit Dimension.* Doubleday. (Background on tacit knowledge.)\n- Stacey, R. D. (1996). *Complexity and Creativity in Organizations.* Berrett-Koehler. (Complexity-science background.)\n- U.S. Army (2012). *Mission Command: Command and Control of Army Forces.* ADP 6-0. (Operational adoption.)\n\nFile v1.0.1:examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md\n\n# Method in Action: Snowden at IBM (1999) and the HBR Synthesis (2007)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nDave Snowden developed the Cynefin framework while leading the Cynefin Centre for Organisational Complexity at IBM's Institute for Knowledge Management (1999-2004). The framework emerged from Snowden's work on knowledge management in complex organizations and his observation that classical management techniques — strategy frameworks, planning methodologies, KPI systems — kept failing when applied to certain kinds of problems. Snowden's diagnosis: those techniques were designed for Complicated problems (knowable cause-effect, expert analysis) and were being misapplied to Complex problems (emergent cause-effect, requires probing).\n\nThe framework had three formative influences:\n\n1. **Knowledge management at IBM.** Snowden observed that knowledge in complex organizations is largely tacit (Polanyi 1966) and contextually bound — it cannot be extracted and codified the way explicit knowledge can. This led to the recognition that knowledge transfer in Complex domains requires different methods than in Complicated.\n\n2. **Complexity science.** The Santa Fe Institute's work on complex adaptive systems (1980s-1990s) gave Snowden the mathematical foundation for the Complex domain: systems with many interacting agents, emergent properties, non-linear dynamics, and irreducible uncertainty.\n\n3. **Practical experience in conflict zones.** Snowden's later application of the framework in the British government's military planning (post-9/11), Singapore's risk-assessment programs, and complex civil-society interventions tested and refined the framework against high-stakes problems.\n\nThe 2007 *Harvard Business Review* paper, co-authored with Mary Boone, was the framework's mainstream introduction. It explicitly placed Cynefin in the management-decision context and provided concrete examples of each domain. The paper's most-cited passages:\n\n> \"Many executives are surprised when previously successful leadership approaches fail in new situations, but different contexts call for different kinds of responses. Before addressing a situation, leaders need to recognize which context governs it — and tailor their actions accordingly. Cynefin, which is Welsh for 'habitat,' encourages leaders to see things from new viewpoints, assimilate complex concepts, and address real-world problems and opportunities. It sorts the issues facing leaders into five contexts defined by the nature of the relationship between cause and effect. Four of these — simple, complicated, complex, and chaotic — require leaders to diagnose situations and act in contextually appropriate ways. The fifth — disorder — applies when it is unclear which of the other four contexts is predominant.\"\n>\n> — Snowden & Boone (2007), p. 69.\n\nThe paper also articulated the framework's most operationally important warning, the **danger of best-practice imposition in the wrong domain**:\n\n> \"The complex domain is where most of the high-stakes strategic decisions of contemporary organizations live. It is also where the most damaging managerial error occurs: the application of complicated-domain methods — exhaustive analysis, expert authority, comprehensive planning — to problems whose underlying dynamics are emergent. The result is a plan that looks impressive in PowerPoint, fails in execution, and generates a post-mortem that blames the execution rather than the underlying mismatch between method and problem.\"\n>\n> — Snowden & Boone (2007), pp. 72-73.\n\nThe paper's most cited example was the response to **Hurricane Katrina (2005)**: the initial response treated the situation as Complicated (apply the standard FEMA playbook), when it was actually Chaotic (the system had collapsed in ways that the playbook did not address). The delay in switching to chaotic-domain method (decisive action to establish order, then sense-respond) cost lives. Subsequent analysis of emergency-response failures in disasters (the COVID-19 pandemic early phase being a recent example) repeatedly shows the same pattern: organizations stuck in Complicated-domain methods when the situation has shifted to Complex or Chaotic.\n\nThe framework has been refined since 2007. Key updates:\n\n- **2014:** \"Simple\" renamed to \"Clear\" to avoid pejorative connotation.\n- **2019:** \"Aporetic\" introduced as a sub-region of Disorder — situations where you don't yet know but the not-knowing is itself the answer.\n- **2020:** \"Liminal\" domains introduced — transitional zones between the main domains, where the situation is moving from one to another.\n\nThe follow-up literature is large:\n\n**Kurtz, C. F., & Snowden, D. J. (2003).** \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. The technical foundation paper, predating the HBR synthesis.\n\n**Snowden, D. J. (2002).** \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111. The knowledge-management foundation.\n\n**Snowden, D. J., & Greenberg, R. (2020).** *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge. The comprehensive 2020 retrospective with updated framework and case studies.\n\n**Mowles, C. (2015).** *Managing in Uncertainty: Complexity and the Paradoxes of Everyday Organizational Life.* Routledge. Academic application to organizational management.\n\n**French, S. (2013).** \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561. The decision-analysis application.\n\nThe framework has shaped operational practice in multiple disciplines:\n\n**Military / defense.** The British and U.S. militaries have integrated Cynefin into officer-training and doctrine development. Counterinsurgency operations (Iraq 2005-2010, Afghanistan throughout) explicitly use the Complex-domain methodology of safe-to-fail experiments at scale. U.S. Army's Mission Command doctrine (2012-) is largely a Cynefin-influenced framework.\n\n**Software development.** Agile methodologies (Scrum, Kanban, Lean) are explicitly the Complex-domain response (Probe-Sense-Respond) to product development. Waterfall is the Complicated-domain method. The literature on \"agile transformation\" failures is largely the story of organizations attempting to import Complicated-domain governance onto Complex-domain work.\n\n**Healthcare.** Treating common illness (Clear), specialized diagnosis (Complicated), chronic-disease management with social determinants (Complex), and pandemic response (Chaotic transitioning to Complex) all require different methods. Healthcare quality movements that ignore the domain differences have produced poor results.\n\n**Public policy.** Wicked problems (climate change, drug policy, inequality, infrastructure) are paradigmatic Complex-domain issues. Treating them with Complicated-domain methods (expert commission, comprehensive plan, single national strategy) is the most-documented policy failure pattern in modern government. Snowden's framework has informed several national-government policy units (UK, Singapore, Australia).\n\n**Organizational change.** Culture change and strategic transformation are emergent, not predictable. Cynefin-influenced change methodologies (Liminal Action, SenseMaker) explicitly use safe-to-fail probes and multi-experiment portfolios instead of single-track transformation plans.\n\n**Crisis management.** First 24 hours of a crisis: Chaotic (Act-Sense-Respond). Next phase: Complex (Probe-Sense-Respond as stabilization continues). Later: Complicated (analyze what happened, build into doctrine). Modern crisis-management protocols increasingly recognize the domain-shift pattern.\n\nThree operational lessons from Cynefin:\n\n**First, the most expensive management error is using a Complicated-domain method on a Complex problem.** Symptoms: detailed plans that fail in execution, post-mortems that blame \"execution\" rather than method, recurrent failure of expert recommendations, \"best practices\" that don't work. The fix: switch to Probe-Sense-Respond, run multiple safe-to-fail experiments, amplify what works.\n\n**Second, the boundary between Clear and Chaotic is the most dangerous.** Long success in a Clear domain (routine operations, mature market, established practice) breeds complacency. A small change (new competitor, regulatory shift, technology disruption) can push the situation into Chaotic without warning. Leaders who respond with \"do what we always do\" preside over the cliff fall. The defense: explicit monitoring for boundary signals; sensors deliberately positioned to catch the transition.\n\n**Third, the framework is diagnostic, not prescriptive.** Cynefin does not tell you what to do; it tells you what *kind* of doing to do. The actual content of each domain's action (which experiments, which experts, which SOP, which decisive crisis intervention) requires domain expertise. Treating Cynefin as a rec\n\nArchive v1.0.0: 5 files, 9881 bytes\n\nFiles: examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md (9098b), references/sources.md (1129b), skill-card.md (2411b), SKILL.md (7002b), _meta.json (126b)","readmeExcerpt":"Skill: Cynefin Owner: deciqai Summary: Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know w... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T17:56:40.866Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/cynefin.json) v1.0.4 | 2026-07-09T11:16:45.631Z | user Refresh: 20","codeSnippets":[],"executableExamples":[{"language":"text","snippet":"Obvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)"},{"language":"text","snippet":"Clear: Sense→Categorize→Respond (SOP/automate) | Complicated: Sense→Analyze→Respond (experts)\nComplex: Probe→Sense→Respond (safe-to-fail experiments, amplify wins)\nChaotic: Act→Sense→Respond (establish order, then re-classify) | Confused: decompose, classify each part"},{"language":"text","snippet":"Domain shifted? (Complicated→Complex from disruption? Clear-Chaotic cliff approaching?)\nClear: deploy SOP; monitor. Complicated: experts; pick defensible alternative.\nComplex: parallel safe-to-fail probes; amplify wins. Chaotic: decisive action; re-diagnose.\nBoundary watch: shift signals | who monitors | re-diagnosis schedule"},{"language":"text","snippet":"Cynefin Diagnosis: \nDomain: [Clear/Complicated/Complex/Chaotic/Confused] | Evidence: [cause-effect, expert agreement]\nMethod: [S-C-R / S-A-R / P-S-R / A-S-R] | Actions: | Mismatch cost (if any):\nBoundary watch: [shift signals | monitoring owner | re-diagnosis schedule]"},{"language":"text","snippet":"Cynefin Diagnosis: Enterprise AI portfolio, 2024–2026\nDomain: Complicated (known model integration) | Complex (agentic reliability) | Chaotic (production incident)\nEvidence: (a) experts converge, best practice transfers, interventions predictable\n (b) experts disagree, behavior emergent/retrospective, estimation fails\n (c) cause-effect in flux, harm compounds with delay\nMethod: (a) S-A-R (b) P-S-R (c) A-S-R\nActions: (a) expert design + evals, ship on plan\n (b) sandbox/shadow, constrained permissions, human-in-loop, capped spend, amplify winning probes\n (c) kill switch, revoke credentials, roll back, contain, then re-diagnose\nMismatch cost: running (b) as (a) → blown timelines + unhandled incident (the \"execution\" post-mortem)\nBoundary watch: model/tool version changes = shift signals | ML-platform team monitors | re-diagnose each release"},{"language":"text","snippet":"Obvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)"}],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: cynefin\ndescription: \"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know what to do first', 'best practice doesn't apply here', 'we need a different approach'. Do NOT activate when: situation is unambiguously routine (execution only); a specialized tool (OODA, expected value) already fits. More: deciqai.com/c/cynefin\"\n---\n\n# Cynefin\n\n## Overview\n\n**Cynefin** (pronounced \"kuh-NEV-in\"; Welsh for \"habitat\") is a sense-making framework by Dave Snowden (IBM, 1999). Its claim: the right decision approach depends on which of five domains the situation falls into — **Clear** (obvious cause-effect, use SOP), **Complicated** (knowable with expertise, use analysis), **Complex** (emergent, probe first), **Chaotic** (absent cause-effect, act first), **Confused** (unknown domain, decompose first). The most common and costly error: treating Complex problems as Complicated.\n\nComposes with `ooda-loop`, `feedback-loops`, `antifragile`, `first-principles`.\n\n## When to Use\n\n- A familiar approach has stopped working and you can't articulate why\n- Experts disagree on the right answer — a crisis unfolding where the previous playbook doesn't apply\n- \"Best practices from X\" imported without checking if the domain matches\n- A team is over-planning something emergent, or \"let's get more data\" when data won't come without action\n- Allocating AI capex or racing AI-native competition: deciding which AI bets are engineering (Complicated), emergent agent/adoption experiments (Complex), or live incidents (Chaotic)\n\n**Not when:** domain is unambiguously Clear (execution only); small-stakes one-shot; specialized framework already fits.\n\n## Coaching Novices (Adaptive Front Door)\n\n**Engine mode:** concrete case → run The Process. **Coach mode:** unfamiliar → guide step by step.\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. Classify the situation (Clear/Complicated/Complex/Chaotic) and match decision method to domain.\n2. Check fit: if unambiguously routine (Clear), skip framework.\n3. Elicit their real case — decision, current method, cause-effect structure.\n> **[WAIT — do not advance until user responds]**\n4. Are cause-effect relationships obvious, knowable, emergent, or absent? Is current method matched?\n> **[WAIT — do not advance until user responds]**\n5. Close: named domain + matched decision method + boundary watch.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Describe:** `Decision/situation: | Current approach: | What worked/not: | Stakeholders:`\n\n### Step 2: Diagnose the domain\n```\nObvious to everyone? (Clear) | Knowable with expertise? (Complicated)\nOnly retrospective? (Complex) | Absent/in flux? (Chaotic) | Unknown? (Confused)\n```\nDiagnostics: 5 experts converge? (Yes → Complicated; No → Complex). Standard best practice works? (Yes → Clear/Complicated; No →"},{"path":"_meta.json","content":"{\n \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n \"slug\": \"cynefin\",\n \"version\": \"1.0.5\",\n \"publishedAt\": 1784224600866\n}"},{"path":"references/sources.md","content":"# Sources — cynefin\n\n> *Primary sources for the [cynefin](../SKILL.md) skill.*\n\n- Snowden, D. J., & Boone, M. E. (2007). \"A leader's framework for decision making.\" *Harvard Business Review*, 85(11), 68-76. The mainstream introduction.\n- Kurtz, C. F., & Snowden, D. J. (2003). \"The new dynamics of strategy: Sense-making in a complex and complicated world.\" *IBM Systems Journal*, 42(3), 462-483. Technical foundation.\n- Snowden, D. J. (2002). \"Complex acts of knowing: Paradox and descriptive self-awareness.\" *Journal of Knowledge Management*, 6(2), 100-111.\n- Snowden, D. J., & Greenberg, R. (2020). *Cynefin: Weaving Sense-Making into the Fabric of Our World.* Cognitive Edge.\n- French, S. (2013). \"Cynefin, statistics and decision analysis.\" *Journal of the Operational Research Society*, 64(4), 547-561.\n- Polanyi, M. (1966). *The Tacit Dimension.* Doubleday. (Background on tacit knowledge.)\n- Stacey, R. D. (1996). *Complexity and Creativity in Organizations.* Berrett-Koehler. (Complexity-science background.)\n- U.S. Army (2012). *Mission Command: Command and Control of Army Forces.* ADP 6-0. (Operational adoption.)\n- Lovell, J., & Kluger, J. (1994). *Lost Moon: The Perilous Voyage of Apollo 13.* Houghton Mifflin. (Apollo 13 example.)\n- NASA (1970). *Report of the Apollo 13 Review Board* (Cortright Report). Washington, DC: NASA. (Apollo 13 example.)\n- Anthropic (2024–2025). \"Building Effective Agents\" and related public agent-engineering guidance. anthropic.com. (2024–2026 AI-decisions example — the \"start simple, add autonomy incrementally, keep humans in the loop, constrain the action space\" posture as Probe–Sense–Respond.)\n- Snowden, D. J., et al. (2020–2021). *Cynefin & Weaving Sense-Making into the Fabric of Our World* / Cynefin.io field guides. Cognitive Edge / The Cynefin Co. (Contemporary restatement of the domains and boundary dynamics used in the 2024–2026 AI example.)"},{"path":"examples/apollo-13-1970-mission-response.md","content":"# Method in Action: The Apollo 13 Mission Response (1970)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nApollo 13 is a compressed, four-day demonstration of every Cynefin domain — and of why survival depended on re-diagnosing the domain at each phase instead of running one method throughout.\n\n**The situation.** On April 13, 1970, roughly 56 hours into a routine lunar mission about 200,000 miles from Earth, oxygen tank 2 in the service module exploded, crippling the spacecraft's power, oxygen, and propulsion. Jim Lovell's report to Houston — \"Houston, we've had a problem\" — marked the instant the mission left its planned domain.\n\n**Phase 1 — Chaotic: Act–Sense–Respond.** In the first hour, cause-effect was absent: telemetry contradicted itself, controllers initially suspected instrumentation failure, oxygen was venting, and fuel cells were dying. There was no playbook for a dying service module en route to the Moon. Flight director Gene Kranz's move was the Chaotic-domain move: act first to establish order, not analyze. Mission Control powered down the command module Odyssey to preserve its re-entry batteries and moved the crew into the lunar module Aquarius as a lifeboat — an improvised, order-establishing action taken before anyone understood the cause. Deliberating would have cost the crew their margin. Only after order was established did the situation become sensible enough to re-classify.\n\n**Phase 2 — Complex: Probe–Sense–Respond.** With the crew stabilized, NASA faced novel problems with no procedures and emergent cause-effect: a lunar module designed to keep two men alive for two days now had to keep three alive for four; carbon dioxide was accumulating because the command module's square lithium hydroxide canisters did not fit the lunar module's round scrubber sockets; the power budget had no precedent. Experts could not simply compute the answer — ground teams probed. Engineers in Houston assembled a canister adapter from only the materials known to be aboard (plastic bags, cardboard, suit hose, tape), tested it, then read the build procedure up to the crew. Simulator teams ran candidate power-down and navigation configurations, sensed what held, amplified what worked, and discarded what failed. This was a portfolio of safe-to-fail experiments run on the ground so that failure would not be fatal in flight.\n\n**Phase 3 — Complicated: Sense–Analyze–Respond.** Some sub-problems were knowable with expertise and moved by analysis: the decision to use a free-return trajectory around the Moon rather than a risky direct abort, the PC+2 engine burn after lunar flyby to speed the return, and the command module power-up sequence — written and verified in simulators by astronaut Ken Mattingly and the ground team before being uplinked. Here the correct move was expert analysis of a knowable system, and NASA used it — but only for the sub-problems that genuinely lived in that domain.\n\n**Phase 4 — back to Clear.** After splashdown on April 17, 197"},{"path":"examples/snowden-at-ibm-1999-and-the-hbr-synthesis-2007.md","content":"# Method in Action: Snowden at IBM (1999) and the HBR Synthesis (2007)\n\n> *Example for the [cynefin](../SKILL.md) skill.*\n\nDave Snowden developed the Cynefin framework while leading the Cynefin Centre for Organisational Complexity at IBM's Institute for Knowledge Management (1999-2004). The framework emerged from Snowden's work on knowledge management in complex organizations and his observation that classical management techniques — strategy frameworks, planning methodologies, KPI systems — kept failing when applied to certain kinds of problems. Snowden's diagnosis: those techniques were designed for Complicated problems (knowable cause-effect, expert analysis) and were being misapplied to Complex problems (emergent cause-effect, requires probing).\n\nThe framework had three formative influences:\n\n1. **Knowledge management at IBM.** Snowden observed that knowledge in complex organizations is largely tacit (Polanyi 1966) and contextually bound — it cannot be extracted and codified the way explicit knowledge can. This led to the recognition that knowledge transfer in Complex domains requires different methods than in Complicated.\n\n2. **Complexity science.** The Santa Fe Institute's work on complex adaptive systems (1980s-1990s) gave Snowden the mathematical foundation for the Complex domain: systems with many interacting agents, emergent properties, non-linear dynamics, and irreducible uncertainty.\n\n3. **Practical experience in conflict zones.** Snowden's later application of the framework in the British government's military planning (post-9/11), Singapore's risk-assessment programs, and complex civil-society interventions tested and refined the framework against high-stakes problems.\n\nThe 2007 *Harvard Business Review* paper, co-authored with Mary Boone, was the framework's mainstream introduction. It explicitly placed Cynefin in the management-decision context and provided concrete examples of each domain. The paper's most-cited passages:\n\n> \"Many executives are surprised when previously successful leadership approaches fail in new situations, but different contexts call for different kinds of responses. Before addressing a situation, leaders need to recognize which context governs it — and tailor their actions accordingly. Cynefin, which is Welsh for 'habitat,' encourages leaders to see things from new viewpoints, assimilate complex concepts, and address real-world problems and opportunities. It sorts the issues facing leaders into five contexts defined by the nature of the relationship between cause and effect. Four of these — simple, complicated, complex, and chaotic — require leaders to diagnose situations and act in contextually appropriate ways. The fifth — disorder — applies when it is unclear which of the other four contexts is predominant.\"\n>\n> — Snowden & Boone (2007), p. 69.\n\nThe paper also articulated the framework's most operationally important warning, the **danger of best-practice imposition in the wrong domain**:\n\n> \"The complex d"}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":"Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know w... Skill: Cynefin Owner: deciqai Summary: Activate when: 'our best practices keep failing', 'experts disagree on the right answer', 'the old playbook isn't working', 'we're in crisis and don't know w... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T17:56:40.866Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/cynefin.json) v1.0.4 | 2026-07-09T11:16:45.631Z | user Refresh: 20","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":1942,"uniquenessScore":52,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-11T01:37:00.444Z","emptyReason":"No screenshots, media assets, or demo links are available."},"primaryImageUrl":null,"mediaAssetCount":0,"assets":[],"demoUrl":null},"ownerResources":{"evidence":{"source":"unclaimed","verified":false,"confidence":"low","updatedAt":"2026-10-11T01:37:00.444Z","emptyReason":"This page has not been claimed by the agent owner."},"hasCustomPage":false,"customPageUpdatedAt":null,"customLinks":[],"structuredLinks":{"docsUrl":null,"demoUrl":null,"supportUrl":null,"pricingUrl":null,"statusUrl":null},"customPage":null},"relatedAgents":{"evidence":{"source":"protocol-neighbors","verified":false,"confidence":"medium","updatedAt":"2026-10-11T03:56:19.051Z","emptyReason":null},"items":[{"id":"8ebccd8e-3863-4187-8355-c3f14e1f9edf","entityType":"agent","canonicalPath":"/agent/iofficeai-aionui","slug":"iofficeai-aionui","name":"AionUi","description":"Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!","url":"https://github.com/iOfficeAI/AionUi","homepage":"https://www.aionui.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-10-09T19:11:12.944Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"id":"b917f68a-ebff-438e-84f8-3f4b2494c0bc","entityType":"agent","canonicalPath":"/agent/activepieces-activepieces","slug":"activepieces-activepieces","name":"activepieces","description":"AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents","url":"https://github.com/activepieces/activepieces","homepage":"https://www.activepieces.com","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-15T02:22:12.426Z","createdAt":"2026-02-25T03:38:12.412Z","downloads":null},{"id":"5cb26759-3a39-483f-94cf-276a98c13bb8","entityType":"agent","canonicalPath":"/agent/cherryhq-cherry-studio","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","url":"https://github.com/CherryHQ/cherry-studio","homepage":"https://cherry-ai.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-11T14:38:40.986Z","createdAt":"2026-02-25T03:38:19.379Z","downloads":null},{"id":"6f6582d0-5d76-4f0f-b81d-86520247950b","entityType":"agent","canonicalPath":"/agent/copilotkit-copilotkit","slug":"copilotkit-copilotkit","name":"CopilotKit","description":"The Frontend for Agents & Generative UI. React + Angular","url":"https://github.com/CopilotKit/CopilotKit","homepage":"https://docs.copilotkit.ai","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-03-25T09:50:57.846Z","createdAt":"2026-02-25T03:39:14.617Z","downloads":null}],"links":{"hub":"/agent","source":"/agent/source/clawhub","protocols":[{"label":"OpenClaw","href":"/agent/protocol/openclew"}]}}}