Clarity Gate
v2.1.3Pre-ingestion verification for epistemic quality in RAG systems. Ensures documents are properly qualified before entering knowledge bases. Produces CGD (Clar...
⭐ 0· 545·0 current·0 all-time
byFrancesco Marinoni Moretto@frmoretto
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name/description (pre-ingestion epistemic checks for RAG) align with the included artifacts: SKILL.md describes a format and validation codes and two small Python utilities implement deterministic claim IDs and document hashing. There are no unrelated binaries, env vars, or config paths requested.
Instruction Scope
SKILL.md instructs the agent to validate and canonicalize CGD documents and references running the bundled scripts to compute IDs and hashes. The instructions operate on local document files (expected for a pre-ingestion verifier) and do not direct data to external endpoints or request unrelated system files. Note: the skill will need access to any documents you ask it to verify (expected behavior).
Install Mechanism
No install spec is provided (instruction-only), and the bundled Python scripts are small reference implementations. Nothing is downloaded or written by an installer during skill setup.
Credentials
The skill declares no required environment variables, no primary credential, and no config paths. The scripts operate on local files and do not read environment secrets in the visible code.
Persistence & Privilege
always is false and the skill is user-invocable with normal autonomous invocation allowed. The skill does not request persistent or elevated system presence, nor does it attempt to modify other skills or global agent settings in the provided materials.
Assessment
This skill appears coherent and implements the stated document-format and hashing behavior locally; it does not request credentials or network access in the visible files. Before installing or enabling: (1) review the full scripts (the provided document_hash.py excerpt is truncated here) to confirm there are no hidden network calls or side effects, (2) run the scripts on non-sensitive test documents or in a sandbox first, and (3) if you will process sensitive documents, ensure the agent only receives the minimal data needed (the skill must read any files you ask it to verify). If you need higher assurance, fetch the upstream repository (listed in SKILL.md) and verify the script hashes or perform a line-by-line review.Like a lobster shell, security has layers — review code before you run it.
latestvk97am9m25bfj9cvkz6qf49vvk9825n4r
License
MIT-0
Free to use, modify, and redistribute. No attribution required.