What makes Guardian Audit notable?

988 downloads. Trust evidence available

How should Guardian Audit be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

Guardian Audit

Tamper-evident audit logger that pairs with Guardian safety skill. Captures safety decisions (halts, approvals, escalations) in an append-only, hash-chained...

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

988

Updated

Oct 11, 2026

Version

1.2.0

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 988 downloads reported by the source. Last updated 10/11/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 11, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 11, 2026
Adoption signal: 988 downloadsadoption · observed Oct 11, 2026
Latest release: 1.2.0release · observed May 25, 2026
Handshake status: UNKNOWNsecurity

What changed

Newest first. Only changes that point at a source are listed.

Release 1.2.0release
- Added top-level compliance documentation: COMPLIANCE-MAPPING.md, LOG-SCHEMA.md, and REPLAY.md moved from 'references/' to the root directory. - Updated skill description to clarify: now only triggers on explicit safety decisions or compliance actions, not general logging requests. - Added a "Privacy & Data Warning" section highlighting best practices and risks, advising caution with sensitive information in logs. - Removed redundant or out-of-date files from the 'references/' directory.
Observed May 25, 2026 · source

Install and run

Setup complexity: low.

clawhub skill install s1732m0wt3pbwh1b2yn4byean986njc4:data-guardian-audit

Install using `clawhub skill install s1732m0wt3pbwh1b2yn4byean986njc4:data-guardian-audit` in an isolated environment before connecting it to live workloads.
No published capability contract is available yet, so validate auth and request/response behavior manually.
Review the upstream CLAWHUB listing at https://clawhub.ai/tooled-app/data-guardian-audit before using production credentials.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-tooled-app-data-guardian-audit/snapshot"

Run-check

$0.02 USD

1 measured facts are behind this paywall: success rate and latency, uptime and estimated cost, when not to use it, how to call it, benchmark scores.

Agents pay $0.02 in USDC. A card payment is $0.50, the smallest a card allows.

Documentation

CLAWHUB

64,369 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: guardian-audit
version: 1.1
description: Tamper-evident audit logger that pairs with Guardian safety skill. Captures safety decisions (halts, approvals, escalations) in an append-only, hash-chained log. Use when Guardian makes a safety decision. Use for compliance audit trails. Does NOT activate on general logging requests like 'log this' or routine queries. Requires explicit safety context or compliance intent to trigger.
Support: [email protected]
---

# Guardian Audit — Tamper-Evident Audit Logger (v1.1)

> *"Trust, but log."*

A companion skill to **Guardian** (or any safety gatekeeper) that captures every decision, action, and escalation in an **append-only, hash-chained audit trail**.

**Why this exists:** Guardian stops bad things. Guardian Audit proves it stopped them — or proves the agent did them anyway.

## Privacy & Data Warning

**Before using Guardian Audit, understand what it captures:**

This skill logs detailed operational data including: file paths, agent reasoning, human escalation responses, system commands, and safety decisions. These logs may contain:
- Internal system paths and configurations
- Sensitive operational details
- Personal or business data processed by agents
- Security decisions and approval context

**Best practices:**
- Review and redact logs before sharing or archiving
- Configure log retention policies appropriate to your compliance requirements
- Restrict log file access to authorized personnel only
- Do not log secrets, passwords, or API keys in agent reasoning fields
- Consider encryption for logs stored on shared or cloud-backed systems

## What It Logs

Every entry includes:

| Field | Purpose |
|-------|---------|
| `timestamp` | ISO-8601 with millisecond precision |
| `sequence` | Monotonic integer, no gaps allowed |
| `previous_hash` | SHA-256 of previous entry (chain integrity) |
| `event_type` | `GUARDIAN_CHECK`, `GUARDIAN_HALT`, `GUARDIAN_APPROVE`, `EXECUTED`, `ESCALATION_RESOLVED` |
| `agent_id` | Anonymous identifier (no platform specifics) |
| `operation` | What the agent tried to do |
| `target` | File / path / database / endpoint |
| `category` | CRITICAL / HIGH / MEDIUM / NON-DESTRUCTIVE |
| `backup_verdict` | VERIFIED / UNVERIFIED / STALE / PARTIAL |
| `backup_checks` | JSON array of which indicators matched |
| `decision` | PROCEED / HALT / AWAITING_HUMAN / DENIED |
| `approver` | `guardian-auto` / `human:anonymous` / `agent:rejected` |
| `agent_reasoning` | The agent's stated justification (quoted) |
| `guardian_notes` | Why Guardian made this decision |
| `outcome` | SUCCESS / FAILURE / TIMEOUT / CANCELLED |
| `entry_hash` | SHA-256 of this entry's content |

## The Hash Chain

```
Entry[N].previous_hash = SHA256(Entry[N-1].content)
Entry[N].entry_hash = SHA256(Entry[N].content)
```

If any entry is modified, every subsequent entry's `previous_hash` fails verification. The chain is self-validating.

## Log Format

Append-only line-delimited JSON (NDJSON), one entry per line:

README.md

# Guardian Audit — Tamper-Evident Audit Logger

> *"Trust, but log."*

**Guardian Audit** is a companion skill to [Guardian](https://github.com/openclaw/guardian) (or any safety gatekeeper) that captures every destructive operation decision in an **append-only, hash-chained audit trail**.

Guardian stops bad things. Guardian Audit proves it stopped them — or proves the agent did them anyway.

---

## Why This Exists

You have a safety gatekeeper. You know when it halts operations. But when something goes wrong, you need proof:

- What did the agent try to do?
- Did the safety tool stop it?
- Did a human approve it anyway?
- What was the agent's reasoning at the time?

Guardian Audit answers these questions with tamper-evident logs.

---

## How It Works

Every safety decision generates a log entry with:

| Field | Example |
|-------|---------|
| `timestamp` | `2026-05-18T14:02:31.847Z` |
| `sequence` | `42` (monotonic, no gaps) |
| `previous_hash` | SHA-256 of previous entry |
| `event_type` | `GUARDIAN_HALT`, `EXECUTED`, `ESCALATION_RESOLVED` |
| `operation` | `rm -rf /tmp/old-builds` |
| `target` | `/tmp/old-builds` |
| `category` | `HIGH` |
| `backup_verdict` | `UNVERIFIED` |
| `decision` | `HALT` |
| `approver` | `guardian-auto` |
| `agent_reasoning` | `"Cleaning up old build artifacts"` |
| `entry_hash` | SHA-256 of this entry |

**Hash chain:** Each entry references the previous. Modify one entry, every subsequent entry's hash fails verification.

---

## Installation

### For OpenClaw

```bash
# macOS / Linux
cp -r guardian-audit/ ~/.openclaw/skills/

# Windows
xcopy /E /I guardian-audit\ %USERPROFILE%\.openclaw\skills\guardian-audit\
```

Add to mandatory skills alongside Guardian.

### Standalone

```bash
pip install -e .
# Or just copy scripts/ anywhere
```

---

## Quick Start

### Log an event manually

```bash
python3 scripts/log-event.py \
  --event-type GUARDIAN_HALT \
  --operation "rm -rf /tmp/old-builds" \
  --target "/tmp/old-builds" \
  --category HIGH \
  --backup-verdict UNVERIFIED \
  --decision HALT \
  --approver guardian-auto \
  --agent-reasoning "Cleaning up old build artifacts" \
  --guardian-notes "No backup coverage detected"
```

### Verify chain integrity

```bash
python3 scripts/verify-chain.py ~/.local/share/guardian-audit/audit.log
# Chain valid: 2 entries, 0 breaks
# Integrity: PASS
```

### Generate compliance report

```bash
python3 scripts/export-report.py audit.log --format markdown
# Generates summary table, event breakdown, integrity statement
```

---

## Log Location

| Platform | Default Path |
|----------|-------------|
| Windows | `%LOCALAPPDATA%\guardian-audit\audit.log` |
| macOS | `~/.local/share/guardian-audit/audit.log` |
| Linux | `~/.local/share/guardian-audit/audit.log` |

---

## What Gets Logged

Guardian Audit captures:
- Every Guardian decision (check, halt, approve)
- Every executed destructive operation
- Every human escalation and resolution
- Every override attempt (if any)

**Does n

_meta.json

{
  "ownerId": "kn77qg2t2rnb458ahv8751shv582rvm6",
  "slug": "data-guardian-audit",
  "version": "1.2.0",
  "publishedAt": 1779716135672
}

COMPLIANCE-MAPPING.md

# Compliance Mapping

How Guardian Audit satisfies regulatory requirements.

## EU AI Act (2026)

| Requirement | Guardian Audit Feature |
|-------------|------------------------|
| Art. 52(1) — High-risk systems log operation | Every tool call logged |
| Art. 52(2) — Automatic logging | Passive, no agent discretion |
| Art. 52(3) — Logs retained | Append-only, human-rotated |
| Art. 52(4) — Access by competent authorities | NDJSON readable without tooling |

## SOC 2 Type II

| Requirement | Guardian Audit Feature |
|-------------|------------------------|
| CC6.1 — Logical access controls | Every access logged with approver |
| CC7.2 — System operations monitored | Real-time event stream |
| CC7.3 — System changes logged | Guardian_HALT/EXECUTED capture changes |
| A1.2 — Availability monitoring | Outcome field tracks failures |

## HIPAA §164.312(b)

| Requirement | Guardian Audit Feature |
|-------------|------------------------|
| Audit controls | All agent activity logged |
| Mechanisms to examine | `scripts/export-report.py` generates readable reports |
| Activity recording | Who (agent_id), what (operation), when (timestamp), result (outcome) |

## GDPR Article 5(1)(d) — Accuracy

Guardian Audit captures the agent's stated reasoning at decision time, enabling post-hoc verification of whether the agent's justification matched reality.

## NIST AI RMF (2024)

| Function | Mapping |
|----------|---------|
| GOVERN 1.1 | Policy enforcement via Guardian + Audit |
| MAP 1.1 | Context documented in guardian_notes |
| MEASURE 2.1 | Metrics from audit trail: halt rate, approval rate, pattern frequency |
| MANAGE 2.1 | Escalation resolution tracked |

LOG-SCHEMA.md

# Log Schema — Guardian Audit

## Entry Fields (all entries)

| Field | Type | Required | Description |
|-------|------|----------|-------------|
| `timestamp` | string (ISO-8601) | Yes | UTC with millisecond precision |
| `sequence` | integer | Yes | Monotonic, starting at 1, no gaps |
| `previous_hash` | string (hex 64) | Yes | SHA-256 of previous entry's `entry_hash` field. First entry uses `"genesis"`. |
| `event_type` | enum | Yes | See Event Types below |
| `agent_id` | string | Yes | Anonymous agent identifier. No platform specifics. |
| `operation` | string | Yes | The tool call or action being logged |
| `target` | string | Yes | File path, database, endpoint, or "N/A" |
| `category` | enum | Yes | CRITICAL / HIGH / MEDIUM / NON-DESTRUCTIVE / N/A |
| `backup_verdict` | enum | No | VERIFIED / UNVERIFIED / STALE / PARTIAL / N/A |
| `backup_checks` | array of objects | No | Which indicators matched |
| `decision` | enum | Yes | PROCEED / HALT / AWAITING_HUMAN / DENIED / LOGGED |
| `approver` | string | Yes | `guardian-auto` / `human:anonymous` / `agent:rejected` / `system` |
| `agent_reasoning` | string | No | Agent's stated justification, quoted verbatim |
| `guardian_notes` | string | No | Why Guardian or safety tool made this decision |
| `outcome` | enum | No | SUCCESS / FAILURE / TIMEOUT / CANCELLED / PENDING |
| `entry_hash` | string (hex 64) | Yes | SHA-256 of this entry's canonical JSON |

## Event Types

| Type | When |
|------|------|
| `GUARDIAN_CHECK` | Guardian scanned operation, backup verified, proceeded |
| `GUARDIAN_HALT` | Guardian halted operation, backup unverified |
| `GUARDIAN_APPROVE` | Guardian auto-approved (backup verified) |
| `EXECUTED` | Operation executed (logged after completion) |
| `ESCALATION_RESOLVED` | Human responded to escalation |
| `MANUAL_APPROVE` | Human manually approved outside Guardian |
| `MANUAL_DENY` | Human manually denied |
| `AGENT_OVERRIDE_ATTEMPT` | Agent tried to override safety (should not happen) |
| `CHAIN_VERIFICATION` | Automated integrity check passed/failed |

## Hash Computation

```python
import json, hashlib

def compute_hash(entry):
    # Canonical JSON: sorted keys, no whitespace
    canonical = json.dumps(entry, sort_keys=True, separators=(',',':'))
    return hashlib.sha256(canonical.encode('utf-8')).hexdigest()

# entry_hash is computed AFTER adding previous_hash but BEFORE adding entry_hash itself
# So entry_hash covers: timestamp, sequence, previous_hash, event_type, ..., outcome
```

## Validation Rules

1. `sequence` must equal previous entry's `sequence` + 1
2. `previous_hash` must equal SHA-256 of previous entry's `entry_hash`
3. `entry_hash` must match recomputed hash of entry content
4. `timestamp` must be >= previous entry's timestamp
5. `timestamp` must not be > 60 seconds in the future
6. `agent_id` must be consistent for a single session (optional but recommended)

## Example Entry

```json
{
  "timestamp": "2026-05-18T14:02:31.847Z",
  "sequence": 42,
  "previ

Github ReposUpdated 3d agoRank 70

AionUi

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!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

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

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.

MCPOPENCLAW

Github ReposUpdated 7mo agoRank 70

CopilotKit

The Frontend for Agents & Generative UI. React + Angular