The Agent Skills Directory

[COMMAND_EXECUTION]: The skill directs the agent to execute a local command-line tool located at ./.planr/tooling/planr. It uses various subcommands such as status show, open, next, ensure-scope, set-checklist, set-blocker, and set-verification to retrieve and manage task states. This execution of local binaries is a core component of the skill's functionality.
[PROMPT_INJECTION]: The skill evaluates information from files like .planr/status/current.json, .planr/plans/*.plan.md, and Git diff outputs. This creates a surface for indirect prompt injection if an attacker can influence the content of these files within the repository.
Ingestion points: The agent reads status JSON files, markdown plan files, and Git diffs from the local repository.
Boundary markers: The instructions do not specify the use of delimiters or explicit directives to ignore embedded instructions within these processed data sources.
Capability inventory: The skill can execute the ./.planr/tooling/planr CLI tool and recommend transitions to other skills like planr-fix or planr-review.
Sanitization: There are no explicit instructions for content sanitization, validation, or escaping of the data ingested from the plan and status files.

planr-status