bmad-agent-architect
Pass
Audited by Gen Agent Trust Hub on Apr 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script during its activation phase using the command
uv run {project-root}/_bmad/scripts/resolve_customization.py. This is used to bootstrap the agent's configuration from the vendor's internal project structure and is a standard operational procedure within this framework. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface by design, as it ingests untrusted data from the project environment.
- Ingestion points:
{project-root}/**/project-context.mdand{project-root}/_bmad/bmm/config.yamlinSKILL.md. - Boundary markers: The skill does not implement explicit delimiters or warnings to ignore instructions embedded within the loaded project files.
- Capability inventory: The agent has the ability to execute shell commands (via
uv run) and invoke other skills (bmad-create-architecture,bmad-check-implementation-readiness). - Sanitization: No sanitization or validation of the ingested file content is performed prior to inclusion in the agent context. This surface is considered a standard feature for technical design agents requiring project context.
Audit Metadata