What makes kimi-delegation-skill notable?

1.5K downloads. Trust evidence available

How should kimi-delegation-skill 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

kimi-delegation-skill

Forces all reasoning and code generation to be delegated to a KIMI (KIMMY) causal language model via HuggingFace Transformers. Use this skill when the agent must never reason or author code itself and must instead proxy all tasks to a KIMI-based model. Skill: kimi-delegation-skill Owner: kn7fss5epqvvjb87r6sn0hk1cd802c8a Summary: Forces all reasoning and code generation to be delegated to a KIMI (KIMMY) causal language model via HuggingFace Transformers. Use this skill when the agent must never reason or author code itself and must instead proxy all tasks to a KIMI-based model. Tags: latest:1.0.0 Version history: v1.0.0 | 2026-02-06T14:35:03.347Z | user - Initial re

OpenClaw

View on ClawHub

Rank

Safety

Downloads

1.5k

Updated

Apr 15, 2026

Version

1.0.0

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1.5K downloads reported by the source. Last updated 4/15/2026.

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 Apr 15, 2026
Protocol compatibility: OpenClawcompatibility · observed Apr 15, 2026
Adoption signal: 1.5K downloadsadoption · observed Apr 15, 2026
Latest release: 1.0.0release · observed Feb 6, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install kn7fss5epqvvjb87r6sn0hk1cd802c8a:kimi-delegation-skill

Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.
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: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-kn7fss5epqvvjb87r6sn0hk1cd802c8a-kimi-delegation-skill/snapshot"

Documentation

CLAWHUB

4,194 characters of source documentation, loaded on request.

Extracted files

4 files captured from the source.

SKILL.md

---
name: kimi-delegation-skill
description: Forces all reasoning and code generation to be delegated to a KIMI (KIMMY) causal language model via HuggingFace Transformers. Use this skill when the agent must never reason or author code itself and must instead proxy all tasks to a KIMI-based model.
license: Proprietary
compatibility: Requires Python, PyTorch, HuggingFace transformers, GPU optional but recommended.
allowed-tools: Python(Read Write)
metadata:
author: internal
version: "1.0"
---
## Purpose

This skill enforces a strict delegation model where the primary agent has zero reasoning or code-authoring authority. All user tasks are forwarded to a KIMI (KIMMY) model loaded via Transformers. The agent acts only as a dispatcher.

## Activation Conditions

Activate this skill whenever:
- The agent must not reason independently.
- All planning, reasoning, and code generation must be authored by a KIMI/KIMMY model.
- Deterministic delegation to an external causal LM is required.

## Execution Steps

1. Initialize `KIMISkill` with a valid local or remote model path.
2. Wrap the `KIMISkill` instance with `Qwen3Coder`.
3. On every user prompt, call `Qwen3Coder.handle_prompt`.
4. The prompt is forwarded verbatim to KIMMY.
5. KIMMY generates the full response.
6. Strip prompt scaffolding and return the result as the final output.

See:
- `scripts/kimi_skill.py`
- `scripts/qwen3_coder.py`

## Inputs and Outputs

**Input:**
A raw user task string.

**Output:**
A dictionary with:
- `author`: Always `"KIMMY"`
- `content`: The generated response with no prompt scaffolding.

## Failure Modes and Edge Cases

- Model path invalid or unavailable: initialization fails.
- Insufficient VRAM: model may fall back to CPU or fail to load.
- Extremely long tasks may exceed context limits.
- If generation fails, no fallback reasoning is permitted.

The agent must not attempt to recover by reasoning itself.

## References

Technical details and architectural rationale are in:
- `references/REFERENCE.md`

assets/README.md

This directory is intentionally empty.

Reserved for future static assets such as configuration templates,
diagrams, or policy documents.

_meta.json

{
  "ownerId": "kn7fss5epqvvjb87r6sn0hk1cd802c8a",
  "slug": "kimi-delegation-skill",
  "version": "1.0.0",
  "publishedAt": 1770388503347
}

references/REFERENCE.md

## Delegation Contract

This skill enforces a non-negotiable authority boundary.

- The agent must never reason.
- The agent must never write or modify code.
- The agent must never post-process output.
- All cognition is performed by KIMMY.

Violating this contract breaks the skill by definition.

## Design Rationale

Using Transformers directly guarantees:
- No hidden orchestration logic
- Full control over generation parameters
- Explicit delegation enforcement

The agent is reduced to I/O plumbing.

## Operational Notes

- Prompt stripping prevents scaffold leakage
- Deterministic low-temperature sampling minimizes variance
- No fallback logic is permitted

Github ReposUpdated 9h 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.