triage

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill performs expected development tasks such as reading local project files and writing summary reports. No indicators of malice, obfuscation, or credential theft were found.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests findings from review files.
  • Ingestion points: Loads data from review files in .claude/plans/.../reviews/.
  • Boundary markers: None identified; the skill parses raw markdown finding content directly.
  • Capability inventory: Limited to interactive user questioning (AskUserQuestion) and writing markdown files to the project directory. No shell command execution or network capabilities were identified.
  • Sanitization: The skill does not explicitly sanitize or validate the content of the review findings before processing.
  • Assessment: The overall risk is negligible because the ingested data is used only for interactive selection and report generation, without triggering sensitive or automated system actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 05:41 PM
Security Audit — agent-trust-hub — triage