知识账本
v1.0.0Evidence-first knowledge auditing skill that upgrades connected knowledge into an auditable conclusion card. It traces which sources support a conclusion, ma...
⭐ 0· 31·0 current·0 all-time
byhaidong@harrylabsj
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (knowledge audit, source traceability, freshness, conflicts, evidence vs inference) match the SKILL.md and reference documents. The skill requires no binaries, env vars, or config paths — all appropriate for an instruction-only auditing skill.
Instruction Scope
Runtime instructions focus on auditing user-supplied material: identify the claim, build a 2–5 source ledger, grade evidence, judge freshness, and surface conflicts. The instructions reference only internal reference files (audit heuristics, conclusion cards) and user-provided documents; they do not direct the agent to read unrelated system files, access external endpoints, or exfiltrate credentials.
Install Mechanism
No install spec and no code files (instruction-only). This minimizes disk writes and arbitrary code execution risk.
Credentials
The skill declares no required environment variables, no primary credential, and no config paths. There are no requests for unrelated secrets or cloud credentials — proportional to the declared purpose.
Persistence & Privilege
always is false (default). The skill is user-invocable and allows model invocation (normal for skills). It does not request persistent system changes or modify other skills' configs.
Assessment
The skill appears coherent and limited to auditing documents you give it. Before using it, avoid pasting highly sensitive secrets or credentials into the inputs — only provide the documents or excerpts you want audited. Note that autonomous invocation is permitted (the default behavior for skills); if you prefer to control when it runs, invoke it manually. If you need more assurance, inspect the reference files included (audit-heuristics.md, conclusion-cards.md) to confirm the exact grading and output format it will use.Like a lobster shell, security has layers — review code before you run it.
conflict-detectionvk977a450pdx8d0bwn9fzcddth9847j0revidence-firstvk977a450pdx8d0bwn9fzcddth9847j0revidence-gradingvk977a450pdx8d0bwn9fzcddth9847j0rfreshness-checkvk977a450pdx8d0bwn9fzcddth9847j0rknowledgevk977a450pdx8d0bwn9fzcddth9847j0rknowledge-auditvk977a450pdx8d0bwn9fzcddth9847j0rlatestvk977a450pdx8d0bwn9fzcddth9847j0rprovenancevk977a450pdx8d0bwn9fzcddth9847j0rresearchvk977a450pdx8d0bwn9fzcddth9847j0rsource-traceabilityvk977a450pdx8d0bwn9fzcddth9847j0rtrustvk977a450pdx8d0bwn9fzcddth9847j0r
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🧾 Clawdis
OSLinux · macOS · Windows