The Agent Skills Directory

[COMMAND_EXECUTION]: The skill directs the agent to execute system-level commands including grep, ls, and git show to reconcile the current state of the repository against planning documents. It also invokes a custom utility named doctrine for scaffolding plans and managing phase lifecycles. These commands are necessary for the skill's primary function of project coordination.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external, potentially untrusted documents which introduces a risk of indirect prompt injection.
Ingestion points: Content is read from slice-nnn.md, design.md, and the results of the review show command.
Boundary markers: The instructions lack specific guidance on using delimiters or protective framing when the agent processes these external files.
Capability inventory: The agent has the ability to read files, perform directory listings, use grep for searching, and execute commands via the doctrine tool.
Sanitization: There are no instructions provided to validate or sanitize the input from design documents before they are used to generate phase criteria.
[DYNAMIC_EXECUTION]: The skill requires the agent to author plan.toml files containing regular expression patterns for the VT-n (Verification Test) mandates. These patterns are interpreted by the doctrine tool at runtime to judge phase completion, representing the generation of executable logic.

plan