bmad-cis-problem-solving
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script
{project-root}/_bmad/scripts/resolve_customization.pyusinguv runat the start and end of the workflow to handle configuration merging and final instructions. - [PROMPT_INJECTION]: The skill processes external data, creating an indirect prompt injection surface:
- Ingestion points: Content is loaded from
solving-methods.csv, project files matched bypersistent_factsglobs, and caller-provided context. - Boundary markers: The instructions lack explicit boundary markers or ignore-rules for the ingested data, instead directing the agent to treat it as 'foundational context'.
- Capability inventory: The agent can execute local commands via
uv runand write output files to the project directory. - Sanitization: No validation or sanitization is performed on the ingested content.
- [EXTERNAL_DOWNLOADS]: The
module-manifest.tomlfile identifies update sources and documentation links targeting the vendor's official GitHub repository and domain (bmad-method.org). These references are documented as standard vendor resources.
Audit Metadata