iterate
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of shell commands and local scripts to manage git state, calculate metrics, and verify work.
- Evidence: The skill executes
git,gh(GitHub CLI),date, andawkcommands, and specifically calls local helper scriptsscripts/loc.shandscripts/snapshot.shwhich perform git index manipulations. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection because it autonomously processes untrusted data (user tasks and repository code) and spawns subagents to review its own changes.
- Ingestion points: The agent reads and processes the user's task and existing repository files (SKILL.md, Operating Posture).
- Boundary markers: While the skill provides internal constraints (Hard Rules), it lacks explicit delimiters or instructions for the agent to ignore potentially malicious embedded instructions within the code it is tasked to edit.
- Capability inventory: The skill is granted capabilities to perform git operations (
git add,git push), interact with the GitHub API via CLI, and execute local shell scripts. - Sanitization: No explicit sanitization or filtering of the repository content or diffs is performed before the content is passed to the primary model or its subagents.
Audit Metadata