The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses uv run to execute a local Python script (resolve_customization.py) located within the vendor's project directory structure to handle configuration merging and finalization logic.\n- [DYNAMIC_EXECUTION]: The workflow allows for the execution of arbitrary actions defined in activation_steps_prepend, activation_steps_append, and on_complete blocks within the project's configuration files (.toml). This creates a mechanism where the agent's behavior is dynamically determined by external configuration content.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from several external sources to ground its analysis and configure its operation.\n
Ingestion points: Processes customize.toml, {skill-name}.toml, {skill-name}.user.toml, config.yaml, and files matched by persistent_facts globs.\n
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are used when loading facts or configuration into the context.\n
Capability inventory: The skill possesses the ability to execute shell commands via uv run and perform filesystem writes to save strategy artifacts.\n
Sanitization: The skill lacks explicit sanitization or validation logic for the content loaded from these external configuration and fact files before it is processed by the agent.

bmad-cis-innovation-strategy