The Agent Skills Directory

[SAFE]: The skill provides standard documentation templates and guidelines. No malicious behavior, obfuscation, or unauthorized data access patterns were identified.
[EXTERNAL_DOWNLOADS]: The reference documentation suggests the use of well-known community tools like @openapitools/openapi-generator-cli and openapi-to-postman for generating SDKs and API collections.
[INDIRECT_PROMPT_INJECTION]: The changelog generation feature represents a data ingestion surface that processes untrusted git commit history.
Ingestion points: Git commit logs and repository history.
Boundary markers: No specific delimiters or safety instructions are defined in the instructions for handling untrusted input.
Capability inventory: Standard repository read and file write access.
Sanitization: The transformation process relies on the agent's ability to summarize and reformat content, which provides some natural filtering but does not explicitly sanitize for instructions.

documentation