regression-suite
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_CONTEXT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external project files including design documents, bug reports, and smoke test reports. Content within these files could potentially influence the agent's behavior during the regression audit. \n
- Ingestion points: Processes design/gdd/systems-index.md, system GDDs, production/qa/bugs/.md, and production/qa/smoke-.md (referenced in Steps 2c and 2d). \n
- Boundary markers: No specific delimiters or instructions to ignore embedded prompts are defined for the ingested files. \n
- Capability inventory: The skill utilizes Bash, Write, and Edit tools to generate manifests and suggest test files. \n
- Sanitization: The skill does not perform sanitization or validation of the ingested Markdown content. \n- [DYNAMIC_CONTEXT_INJECTION]: The skill uses the
!command`` syntax at the top of SKILL.md to execute a local shell script (yaml-helper.sh) at load time:bash "${CLAUDE_SKILL_DIR}/../../hooks/yaml-helper.sh" resolve_config --keys automation,workflow,qa.level,system_overrides. This command is used to resolve environment-specific configuration keys. The script is located in a relative path within the project's directory structure and represents legitimate project-specific tooling.
Audit Metadata