The Agent Skills Directory

[SAFE]: The skill defines a logical and safe workflow for planning, using templates and rules to ensure tasks are well-defined and small.
[COMMAND_EXECUTION]: The skill utilizes the bd (Beads) command-line interface for task management (creation and updates). These operations are restricted by the allowed-tools frontmatter and represent the primary intended functionality of the skill.
[INDIRECT_PROMPT_INJECTION]: This skill possesses an attack surface as it processes architecture documents (ARCH-*.md) from the file system. A maliciously crafted architecture document could attempt to influence the agent's task creation process. However, the use of a strict template and explicit 'Done when' criteria serves as a natural boundary, and the impact is limited to the management of development tasks within the project scope.
[DATA_EXPOSURE_AND_EXFILTRATION]: No patterns for sensitive data access, credential harvesting, or unauthorized network communications were found.

pm-planning