functional-discovery

To find existing community skills, use the functional discovery agent during planning to query public registries and local installations for overlapping functionality, filtering duplicates to prevent redundant development.

Duplicate agent detection works by querying multiple public registries in parallel with timeout handling, then deduplicating the results across sources while applying a tiered trust model that prefers higher-trust artifacts.

Yes, safe installation is supported by checking for already-installed artifacts, validating frontmatter and file size, adding provenance metadata, and requiring explicit user approval before writing any local files.

The best way to avoid rebuilding existing solutions is running functional discovery with a concise feature description to locate verified community tools that match your planned functionality before starting development.

Trust-tier filtering prevents unreliable community agents by applying a tiered trust model during deduplication, preferring higher-trust artifacts across queried registries to ensure only verified tools are suggested.

When registry queries time out during community tool discovery, the Skill handles errors gracefully to ensure the workflow continues, gathering available candidate artifacts and deduplicating successful responses.