The Agent Skills Directory

[SAFE]: The skill logic and tool usage (Bash, Read, Write, Edit) are consistent with its purpose as a code implementation agent. No suspicious network activities, credential access, or persistence mechanisms were identified.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes artifacts (PRDs, checklists) that may contain attacker-controlled instructions.
Ingestion points: Project artifacts under .dev/<topic-slug>/ such as prd.md, checklist.md, and context-notes.md.
Boundary markers: None; the instructions do not specify the use of delimiters or 'ignore' warnings for the data ingested from files.
Capability inventory: Significant capability surface including Bash command execution, file modification (Edit, Write), and autonomous subagent creation (Task).
Sanitization: None; the skill is instructed to adopt tasks and requirements from external files directly into its execution loop.

implement