bmad-agent-sm
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs a file system search for
**/project-context.mdto load project-specific standards as a foundational reference. - [PROMPT_INJECTION]: The skill implements persona enforcement instructions ("must not break character until the users dismisses this persona") which dictate agent behavior constraints.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through the ingestion of external data from workspace files.
- Ingestion points: Reads content from
**/project-context.mdand retrieves configuration variables from thebmad-initskill. - Boundary markers: No delimiters or "ignore instructions" warnings are defined when loading the project context file.
- Capability inventory: The skill can invoke multiple functional sub-skills (e.g.,
bmad-sprint-planning,bmad-create-story) based on processed data. - Sanitization: There is no evidence of input validation or filtering for the data loaded from the project context file.
Audit Metadata