The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the standard shell mv command to manage the lifecycle of plan files, moving them between docs/plans/new/, docs/plans/, and docs/plans/done/ directories as tasks progress. This is consistent with its stated purpose of document management and process orchestration.
[INDIRECT_PROMPT_INJECTION]: The workflow is designed to ingest and process data from external sources, such as 'handoff summaries' from other models (AW-10) and 'rework briefs' (AW-17).
Ingestion points: Data enters the context via external handoff summaries and user-provided rework briefs (documented in SKILL.md).
Boundary markers: The skill includes Rule AW-16 ('An external review is evidence, never a substitute for your own verification'), which explicitly mandates first-party verification of all external content before adoption.
Capability inventory: The workflow has access to file-system operations (via mv) and other project-management tools mentioned in the integration guide.
Sanitization: The requirement for manual verification of external content serves as a procedural sanitization layer against untrusted instructions.
[EXTERNAL_DOWNLOADS]: The documentation references external projects 'superpowers' and 'graphify' hosted on GitHub. These are informational links for the user and do not involve automated execution of remote scripts or unauthorized package installations.

agentic-workflow