onboard
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and summarizes content from local project files (source code, design documents, tests) which could theoretically contain malicious instructions if the repository is compromised. However, the skill lacks network access and requires user approval for file writes, limiting the potential impact.
- Ingestion points: The skill processes files from src/, design/, tests/, production/, CLAUDE.md, and .claude/agents/ (SKILL.md).
- Boundary markers: There are no specific delimiters or instructions used to separate untrusted project content from the agent's core instructions.
- Capability inventory: The skill employs Read, Glob, Grep, and Write tools (SKILL.md).
- Sanitization: Project content is interpolated directly into the summary template without specific escaping or filtering.
Audit Metadata