bmad-tea-trace
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches vendor-specific tools and dependencies from the official bmad-code-org repositories and the NPM registry using npx (e.g., bmad-method-test-architecture-enterprise).
- [COMMAND_EXECUTION]: The skill executes local Python scripts residing within the project's _bmad/scripts/ directory using uv run to resolve customizations and configurations. It also invokes CLI tools for migration and skill management via npx.
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect injection by ingesting untrusted external data.
- Ingestion points: The skill reads external data from live-verification-results.json, project source code, and requirement documents (story/epic files).
- Boundary markers: While the skill uses structured headers and backticks for paths, it lacks explicit safety instructions for the agent to ignore instructions embedded within the ingested test evidence or source code.
- Capability inventory: The skill possesses the ability to write files to the project directory and execute shell commands through the CLI toolchain.
- Sanitization: The workflow performs JSON and YAML schema validation, including major version checks for live results, to ensure data integrity during processing.
- [DYNAMIC_EXECUTION]: The skill instructions include significant blocks of JavaScript logic for the agent to execute at runtime. This logic is used for complex data rollups, git SHA prefix matching for evidence freshness, and deterministic coverage calculations.
Audit Metadata