The Agent Skills Directory

[COMMAND_EXECUTION]: The skill explicitly instructs the agent to use shell commands (e.g., ls docs/monorail/) to verify the existence of project files and to count directories for automated slug generation. This is used to bypass the limitations of search tools that may skip gitignored paths.
[DYNAMIC_EXECUTION]: The 'Prototype' resolution mode in map-mode.md directs the agent to generate and run 'small throwaway runnables' or UI variants to test logic or behavior. While intended for testing 'behavior feel', this involves the dynamic generation and execution of code within the agent's environment.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect injection by processing untrusted data.
Ingestion points: The skill ingests data from 'primary sources' (official documentation, source code, specifications, and first-party APIs) when resolving research tickets in Map mode.
Boundary markers: No specific delimiters or instructions to ignore embedded commands within the ingested research data are specified.
Capability inventory: The skill has capabilities for shell execution (ls), file writing (docs/monorail/), and executing generated prototype scripts.
Sanitization: No explicit sanitization or validation of the content retrieved from primary sources is mentioned before it is processed or used to form answers.

rail-align