handing-off

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses basic shell commands (mkdir, date) to manage directory structures and file naming for handoff reports. These operations are restricted within the skill's defined configuration and serve the primary purpose of local file management.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests data from external sources like pull requests, issues, and commit messages to generate summaries.
  • Ingestion points: Pull requests, issues, and file content referenced in SKILL.md.
  • Boundary markers: Absent; however, the instructions emphasize summarizing metadata and 'why' rather than verbatim content duplication.
  • Capability inventory: Uses the Write tool to save reports in .claude/reports/ (specified in SKILL.md).
  • Sanitization: No specific sanitization or escaping of external content is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 05:18 AM
Security Audit — agent-trust-hub — handing-off