eod
Pass
Audited by Gen Agent Trust Hub on Jul 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands using
gitand text-processing utilities (sed,awk,grep) to extract repository data and commit history. These are standard operations for the skill's intended purpose and are executed within the agent's shell environment.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes git commit messages (subjects and bodies), which are untrusted external data that could contain malicious instructions designed to influence the LLM summary.\n - Ingestion points: Git commit history retrieved via
git logcommands inSKILL.md.\n - Boundary markers: Absent; the commit content is interpolated into the prompt without explicit delimiters or instructions to the model to ignore embedded commands.\n
- Capability inventory: Local file system read access and shell execution for
gitcommands.\n - Sanitization: Commit messages are truncated to 300 characters, which limits the volume of data but does not filter for malicious content.
Audit Metadata