The Agent Skills Directory

[SAFE]: The skill focuses on standard project management and documentation tasks. It manages configuration files (plan.toml) and design rationales (plan.md) within the local workspace without attempting network access or credential harvesting.\n- [COMMAND_EXECUTION]: The skill uses local read-only commands such as grep, ls, and git show to verify file existence and content consistency. This is a legitimate safety check to prevent the agent from proceeding with stale design premises.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting external data from project files.\n
Ingestion points: The skill reads untrusted data from slice-nnn.md and design.md.\n
Boundary markers: No specific delimiters or instructions are provided to the agent to ignore potential commands embedded in these files.\n
Capability inventory: The agent can execute the doctrine CLI and local system utilities like grep and git.\n
Sanitization: There is no evidence of sanitization or validation of the content ingested from the design and slice documents.

plan