codebase-recon
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses standard shell commands including
git,grep,sort,uniq, andwcto perform local codebase analysis. These operations are limited to accessing repository history and metadata. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from git history during the reconnaissance phase.
- Ingestion points: Git metadata such as commit messages, author names, and filenames extracted via
git logandgit shortlogin SKILL.md. - Boundary markers: The skill does not provide explicit delimiters or instructions to the agent to treat git log output as untrusted content.
- Capability inventory: The agent utilizes shell command execution for analysis and has the capability to write the final report to a local markdown file.
- Sanitization: There is no explicit sanitization or escaping of git log content before it is summarized into the final report.
Audit Metadata