write-flow-guide
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bundled Python script (
scripts/flow-guide) to manage guide state, render HTML reports, and start or stop a local review server. It also coordinates with verification sub-skills such asvet-flow-guideandrun-flow-guide. - [INDIRECT_PROMPT_INJECTION]: The skill presents an indirect prompt injection surface due to its requirement to ingest and analyze untrusted external project files and application source code.
- Ingestion points: The skill reads
requirements.md,design.md,tasks.md, and project source code to derive test cases. - Boundary markers: The instructions do not specify explicit delimiters or "ignore instructions" markers for the ingested content.
- Capability inventory: Capabilities include writing to the
.skills/directory, executing the internal CLI tool, and invoking further sub-skills. - Sanitization: The internal script uses standard JSON handling and provides HTML escaping when generating the rendered flow guide.
Audit Metadata