gsd-ingest-docs
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from documentation files (ADRs, PRDs, SPECs) within a repository.
- Ingestion points: Discovers and reads files from directory paths like
docs/adr/,docs/prd/,docs/specs/, and root-level markdown files matching{ADR,PRD,SPEC,RFC}-*.md. - Boundary markers: Mentions 'approval gates' and a 'BLOCKER safety rule' within the execution context, but defines no specific delimiters or markers to isolate ingested content from the agent's internal instructions.
- Capability inventory: The skill is permitted to use
Write,Edit,Bash, andAgenttools, which allow it to modify repository files and execute commands based on synthesized content. - Sanitization: No explicit sanitization, content filtering, or escaping of the ingested document data is described in the prompt.
Audit Metadata