The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill requires the agent to ingest and process files from the local repository, including configuration, shared modules, and existing tests.\n
Ingestion points: SKILL.md directs the agent to read repository guidance, momentic.config.yaml, and nearby modules.\n
Boundary markers: The instructions do not define specific delimiters or "ignore" markers for the ingested data.\n
Capability inventory: The skill allows the agent to modify the file system and execute repository tools.\n
Sanitization: No sanitization or data validation methods are specified for the processed repository content.\n- [COMMAND_EXECUTION]: The skill provides instructions for the agent to run development-related commands such as linters, typecheckers, and test runners.\n
Evidence: SKILL.md contains instructions for executing "fast repository checks" and "Momentic tests".\n
Control Mechanism: The skill includes a requirement for the agent to obtain explicit user confirmation before running any end-to-end tests.\n- [DYNAMIC_EXECUTION]: The skill instructs the agent to create and run scripts for setting up deterministic test environments.\n
Evidence: SKILL.md encourages the building and execution of "idempotent endpoints or scripts that seed deterministic records and state".

momentic-spec