recap-lite

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to retrieve and display the current date, git status, and the last three git log entries. These are standard operations for development workflows.\n
  • Evidence: date "+šŸ• %H:%M %Z (%A %d %B %Y)" && git status --short && echo "---" && git log --oneline -3 in SKILL.md\n- [COMMAND_EXECUTION]: The skill uses filesystem commands to identify and list the most recent handoff file from a specific directory.\n
  • Evidence: ls -t ψ/inbox/handoff/*.md 2>/dev/null | head -1 in SKILL.md\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external file content.\n
  • Ingestion points: Files matching the pattern ψ/inbox/handoff/*.md are read into the agent context (SKILL.md).\n
  • Boundary markers: Absent. There are no explicit delimiters or instructions to the agent to treat the file content as untrusted data.\n
  • Capability inventory: The skill performs read-only operations and displays the summary; it lacks network or write capabilities that could be exploited across its instructions.\n
  • Sanitization: Absent. No filtering or validation is performed on the content of the ingested files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 05:15 PM
Security Audit — agent-trust-hub — recap-lite