The Agent Skills Directory

[SAFE]: The skill implements a read-only document review process for project-specific change and delta files.\n- [SAFE]: No network operations, external downloads, or shell command executions were identified.\n- [SAFE]: Metadata and instructions are consistent with the stated purpose of a quality-of-handoff review tool.\n- [SAFE]: No obfuscation, persistence mechanisms, or privilege escalation patterns were found.\n- [INDIRECT_PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection through the processing of untrusted thread documents.\n
Ingestion points: The agent reads change.md, documents in the delta/ folder, log.md, and seed.md.\n
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present for the ingested data.\n
Capability inventory: The skill is restricted to reading project documentation and writing findings to the local .pending-reviews/ directory. No network operations, shell command execution, or dynamic code evaluation capabilities were detected.\n
Sanitization: No explicit sanitization or validation of the content within the ingested documents is performed before the review logic is applied.

review-change