lgnh-review-conflict
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local git and shell commands to inspect repository metadata and current state.
- Evidence: Instructions include running
git status,git log,git diff, andcaton files within the.gitdirectory (such asMERGE_HEADandrebase-merge/onto). - [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes untrusted data from commit logs and code changes.
- Ingestion points: Reads external content from git history and resolved file diffs in SKILL.md.
- Boundary markers: Absent; the skill does not wrap processed code or commit messages in delimiters to distinguish them from instructions.
- Capability inventory: Restricted to read-only git and shell operations; no network access or file-writing tools are utilized.
- Sanitization: None; the content of the diffs and logs is analyzed as-is without validation or filtering.
Audit Metadata