The Agent Skills Directory

[SAFE]: The skill correctly implements technical document review logic by utilizing tools like Grep, Read, and Bash to verify documentation against project code. No hardcoded credentials, malicious downloads, or unauthorized network operations were identified.- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection because it processes content from external technical specifications. A malicious specification file could attempt to influence the agent's behavior during the review process.
Ingestion points: The document content is read into the agent's context using the cat command within the mcp__codex__codex prompt.
Boundary markers: The prompt lacks explicit delimiters or instructions for the model to ignore embedded commands within the read document content.
Capability inventory: The skill has access to file system tools (Read, Grep, Glob, Bash) and the ability to dispatch sub-tasks via mcp__codex__codex.
Sanitization: No escaping or sanitization is performed on the content read from technical specifications before it is processed by the model.

review-spec