spec-grill
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local helper scripts using
nodeand callsgit logto extract repository history. These operations are limited to the skill's own scripts and the target repository provided by the user. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from the target repository to generate its reports and specifications.
- Ingestion points: Files including
README.md,spec/charter.md,spec/system-map.md,CLAUDE.md, and git commit messages are read inscripts/extract-signals-core.js. - Boundary markers: The skill uses internal grouping and deterministic ordering but does not explicitly wrap all external inputs in safety delimiters during the signal extraction phase.
- Capability inventory: The skill has the ability to read from the filesystem, write the
spec/capabilities.mdfile, and executegitcommands. - Sanitization: Identifiers extracted from the repo are passed through
slugifyCandidateinscripts/extract-signals-shared.jsto remove non-alphanumeric characters.
Audit Metadata