ps-show-me-your-work
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from agent transcripts and user-provided strings which may contain untrusted content. 1. Ingestion points: Reads from agent-transcripts/ directory and accepts command-line arguments via scripts/log.sh. 2. Boundary markers: Instructions explicitly restrict file access to the active workspace and warn against accessing unrelated project directories. 3. Capability inventory: The skill executes local shell scripts and writes to TSV and HTML files. 4. Sanitization: The scripts/log.sh script includes a clean() function that strips control characters and prefixes formula-triggering characters (=, +, -, @) with a single quote to prevent spreadsheet injection.
- [COMMAND_EXECUTION]: The skill invokes local shell scripts (scripts/log.sh) and Python utilities (scripts/render-report.py) for logging and report rendering. These commands are scoped to the skill's own directory and the current workspace.
Audit Metadata