Smart ClawdBot documentation access with local search index, cached snippets, and on-demand fetch. Token-efficient and freshness-aware.
Security Analysis
high confidenceThe skill's instructions, required artifacts, and behavior are consistent with a local documentation-access helper; it reads local ClawdBot doc caches and uses the platform web_fetch tool to retrieve pages when needed.
The name/description (local search index, cached snippets, on-demand fetch) match the instructions: checking ~/clawd/data/ for snippets/index/cache and using web_fetch to fetch docs.clawd.bot. Minor inconsistency: the SKILL.md header/version metadata lists v2.2.0 while the Version Info table shows 2.1.0 — likely a small doc mismatch but not a functional red flag.
Instructions explicitly tell the agent to list and cat files under ~/clawd/data/ (docs-snippets, docs-index.json, docs-cache) and to call web_fetch to fetch docs.clawd.bot. This is coherent for a local doc-helper, but it does require filesystem reads in the user's home directory and will access any files present there. The SKILL.md refers to a dependency (web_fetch) but the registry metadata did not declare required binaries; the agent should ensure web_fetch exists before relying on it.
No install spec and no code files (instruction-only). This is low-risk: nothing is downloaded or written by an installer step.
The skill declares no environment variables or credentials and does not request external secrets. It does read local files (index, snippets, docs-stats.json). Reading docs-stats.json (usage tracking) is reasonable for this skill but may contain telemetry; the skill does not request unrelated credentials or system config.
always is false and the skill does not request system-wide configuration changes or permanent presence. It only describes reading local caches and invoking web_fetch when needed.
Guidance
This skill appears coherent and designed to read local ClawdBot documentation caches and fetch pages from docs.clawd.bot when needed. Before installing: 1) confirm your agent runtime grants web_fetch and that web_fetch is trusted/available, 2) verify the ~/clawd/data/ directory only contains documentation you are comfortable exposing to the agent (docs-stats.json may contain usage/telemetry), and 3) if you want stricter limits, restrict the agent's filesystem access or remove sensitive files from ~/clawd/data/ so the skill can only read intended documentation. The version number mismatch in the SKILL.md is minor but worth noting.
Latest Release
v2.2.0
MAJOR: Migrated from brightdata to native web_fetch tool - zero cost documentation access! Removed mcporter dependency. Added curl-based refresh script.
More by @aranej
Published by @aranej on ClawHub
