plan-deep-review
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection (Category 8). 1. Ingestion points: It reads CLAUDE.md, TODOS.md, and project files using Read, Glob, and Grep tools. 2. Boundary markers: The instructions lack delimiters or explicit warnings to the agent to ignore instructions embedded within the files being reviewed. 3. Capability inventory: The agent has access to the Bash tool, allowing for shell command execution based on context. 4. Sanitization: Content from external files is interpolated into the context without specific filtering.
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform repository audits using git log, git diff, and the GitHub CLI (gh). These commands are used appropriately to gather context for the review process and do not include dangerous or unauthorized operations.
- [SAFE]: No evidence was found of obfuscated code, hardcoded credentials, malicious persistence mechanisms, or unauthorized data exfiltration to external domains. The frontmatter properly limits tool access and disables model invocation where appropriate.
Audit Metadata