The Agent Skills Directory

[COMMAND_EXECUTION]: The activation sequence in SKILL.md executes a Python script located at {project-root}/_bmad/scripts/resolve_customization.py to merge configuration files during setup.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by automatically loading files matching **/project-context.md into the agent's persistent memory for context. * Ingestion points: SKILL.md (Step 4) and customize.toml (persistent_facts array). * Boundary markers: No delimiters or sanitization warnings are applied to the ingested content from these files. * Capability inventory: The agent is instructed to execute local Python scripts and utilizes subprocesses for tasks like web research as described in write-document.md. * Sanitization: There is no evidence of content filtering or validation for the ingested project facts.
[EXTERNAL_DOWNLOADS]: The instruction set in write-document.md suggests the use of subprocesses for web searching and research tasks, which establishes a capability for outbound network access during the research phase.

bmad-agent-tech-writer