arkweb-security-patch-judge
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's instructions are focused entirely on domain-specific logic for managing security patches within the ArkWeb project. It enforces strict JSON output formats and explicit state transition rules.
- [DATA_INGESTION]: The skill is designed to ingest and evaluate data from external sources, including patch files (.patch, .diff) and build logs (SKILL.md, Rule 8).
- Ingestion points: Processes content from 02_patch_fetch.json and associated patch files, as well as build verification outputs.
- Boundary markers: The skill does not define explicit delimiters or 'ignore' instructions for the content of the files it reads.
- Capability inventory: The skill produces structured JSON output used for workflow routing. It does not execute shell commands or perform network requests itself.
- Sanitization: No explicit sanitization or validation of the patch content is performed within the prompt instructions.
Audit Metadata