gsd-eval-review
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from workspace files (such as
AI-SPEC.md) and user-provided arguments to perform its audit. It lacks explicit boundary markers or sanitization for this external data. While performing the audit, the agent uses powerful tools includingBash,Write, andAgentcalls, which creates a potential surface for instructions hidden within project files to influence the agent's behavior. - Ingestion points: Reads
AI-SPEC.mdand references local workflow files in~/.claude/gsd-core/. - Boundary markers: None present in the instructions to delimit untrusted file content.
- Capability inventory:
Read,Write,Bash,Glob,Grep,Agent,AskUserQuestion. - Sanitization: No explicit sanitization or validation of the ingested file content or phase arguments is defined.
Audit Metadata