The Agent Skills Directory

[SAFE]: The skill is a comprehensive documentation set for the @upstash/redis library and its related search and rate-limiting extensions. It contains technical guides, installation instructions, and API examples consistent with its stated purpose.
[SAFE]: The documentation provides clear guidance on secure credential management, explicitly warning against hardcoding secrets and recommending the use of environment variables or .env files.
[SAFE]: All external references, including package names (e.g., @upstash/redis, @upstash/ratelimit) and URLs (e.g., upstash.com, github.com/upstash/redis-js), belong to the official vendor or well-known services.
[INDIRECT_PROMPT_INJECTION]: As a database client, the skill exposes a surface for indirect prompt injection by facilitating the retrieval of potentially untrusted data from Redis into the agent context.
Ingestion points: Data is retrieved from Redis through various methods described in the data structure guides, such as redis.get, redis.hgetall, redis.json.get, and search index queries.
Boundary markers: The skill documentation does not provide explicit examples of using boundary markers or delimiting instructions for data retrieved from the database.
Capability inventory: The skill is focused on Redis network operations and does not include dangerous capabilities such as local subprocess execution, file system writes, or persistence mechanisms.
Sanitization: The SDK implements automatic serialization and deserialization for native JavaScript types (numbers, booleans, objects), reducing the need for manual parsing of string-based data.

upstash-redis-js