report-qa

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection due to its heavy reliance on external project data files.
  • Ingestion points: The skill instructions in SKILL.md direct the agent to cross-reference report content with specs-quality.json, plans-spatial.json, punch-list.json, change-order-log.json, procurement-log.json, quality-data.json, safety-log.json, schedule.json, and project-config.json.
  • Boundary markers: The skill lacks explicit instructions to treat the data in these JSON files as untrusted or to ignore any embedded instructions or prompts found within them.
  • Capability inventory: The agent is instructed to modify report data and regenerate report sections based on the results of the QA checks. It does not appear to have access to network tools or arbitrary shell execution.
  • Sanitization: There is no mention of sanitizing or validating the contents of the ingested project files before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 11:09 AM
Security Audit — agent-trust-hub — report-qa