The Agent Skills Directory

[SAFE]: No security issues were detected. The skill provides legitimate documentation and integration examples for a well-known managed service using official libraries and best practices.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates vector database RAG (Retrieval-Augmented Generation) patterns, which are inherently subject to indirect prompt injection if the ingested data is untrusted.
Ingestion points: Data enters the system via index.upsert and index.query methods as documented in SKILL.md and references/deployment.md.
Boundary markers: None explicitly implemented in the example snippets; the skill provides the interface rather than a complete application logic.
Capability inventory: The skill enables data storage and retrieval via the Pinecone API; it does not grant the agent high-privilege system access or arbitrary network capabilities beyond the service.
Sanitization: Standard database operations are shown without specific content sanitization, which is expected for this type of technical integration guide.

pinecone