inspection-tracker
Pass
Audited by Gen Agent Trust Hub on Mar 17, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE]: No security threats, malicious code patterns, or safety guideline bypass attempts were identified in the skill's instructions.
- [NO_CODE]: The skill consists of markdown documentation and data models without any executable scripts or binary files, minimizing the attack surface.
- [PROMPT_INJECTION]: The skill functions by ingesting external data from files like specs-quality.json, schedule.json, and plans-spatial.json, which creates an indirect prompt injection surface. 1. Ingestion points: specs-quality.json, schedule.json, and plans-spatial.json (SKILL.md). 2. Boundary markers: None identified; data is directly interpolated into inspection records and intelligence queries. 3. Capability inventory: Local file read and write operations restricted to logging (inspection-log.json, safety-log.json) and report generation (SKILL.md, references/skill-detail.md). 4. Sanitization: None identified; the skill relies on the integrity of the project's internal data files.
Audit Metadata