The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using uv run to invoke a customization resolver script (resolve_customization.py) located in the project's _bmad/scripts/ directory. This occurs during skill activation and validation steps.
[COMMAND_EXECUTION]: The workflow involves running playwright-cli to perform live browser-based evidence collection, such as monitoring network traffic and capturing screenshots of target URLs.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a significant attack surface for indirect prompt injection because its core purpose is to ingest and analyze untrusted external data.
Ingestion points: The skill reads project-level artifacts including tech-spec.md, PRD.md, story files, and various evidence sources like logs and test results (e.g., in steps-c/step-01-load-context.md and steps-c/step-03-gather-evidence.md).
Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" warnings when processing the content of these external files.
Capability inventory: The agent has access to file system tools (read_file, write_file, list_files, search_repo, glob) and command execution capabilities (uv, playwright-cli).
Sanitization: There is no evidence of sanitization, validation, or filtering of the content retrieved from the audited documents before it is processed by the agent.

bmad-testarch-nfr