The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes a repository-local CLI tool located at ./.planr/tooling/planr to initialize projects and scaffold new planning files.
[PROMPT_INJECTION]: The skill demonstrates an attack surface for indirect prompt injection by design, as it must process untrusted external data (bug reports, task documents, review findings) to generate its output.
Ingestion points: Processes user requests, task documents, bug reports, and review findings as identified in SKILL.md.
Boundary markers: No specific delimiters or "ignore instructions" markers are used when interpolating external content.
Capability inventory: The agent can write files to the .planr/ directory and execute the ./.planr/tooling/planr local binary.
Sanitization: The instructions advise the agent to "rewrite it into an execution contract instead of copying the original text verbatim," which serves as a manual abstraction layer but does not constitute technical sanitization of the input.

planr-plan