quick-context
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local project files that are then used to set constraints and priorities for the agent.
- Ingestion points: The skill reads
CLAUDE.md,.claude/CLAUDE.md,CLAUDE.local.md, and any files referenced via@...imports from a user-provided or auto-resolved directory (SKILL.md). - Boundary markers: There are no boundary markers or instructions to treat the content of these files as untrusted; the agent is explicitly told to "Internalize the project's purpose, architecture, commands, conventions, constraints, priorities, and warnings" (SKILL.md).
- Capability inventory: The skill enables file-reading capabilities to ingest markdown context.
- Sanitization: No sanitization or validation of the ingested file content is performed before the agent internalizes the instructions.
Audit Metadata