The Agent Skills Directory

[SAFE]: The skill is a documentation-only asset providing instructional guidance for reviewing implementation plans. It does not include any scripts, binaries, or executable code.
[SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were detected.
[SAFE]: Indirect Prompt Injection Risk Assessment: (1) Ingestion points: The skill processes implementation plans from PLAN.md or the output of plan_read. (2) Boundary markers: No specific delimiters are defined to isolate untrusted plan content from the agent's instructions. (3) Capability inventory: The skill does not perform subprocess calls, file writes, network operations, or any form of code execution. (4) Sanitization: No sanitization or validation of input content is performed. Conclusion: While an attack surface exists for indirect prompt injection, the lack of exploitable capabilities renders the risk negligible.

plan-review