OpenAPI 2 Skill – progressive disclosure of OpenAPI specs for agents
Progressive disclosure beats MCP tool flooding for large OpenAPI specs.

Just docs and OpenAPI spec—no actual agent code visible anywhere.
Customer support teams, SaaS companies
Intercom Fin · Zendesk AI · Help Scout
Progressive disclosure beats MCP tool flooding for large OpenAPI specs.
It extracts focused, executable operations from giant OpenAPI files (the GitHub REST YAML is shown) to shrink context and avoid sidecar adapter sprawl — a pragmatic answer to token bloat and brittle ad-hoc integrations. Useful and concrete: if it actually generates tidy, updateable skill units and runtime hooks it saves a lot of maintenance. That said, the idea competes with existing LangChain/openai-function patterns; the repo will need clear runtime, versioning, and update strategies to feel like more than a nicer converter.
Single source of truth beats drifting Postman collections, but early alpha.
Spec document ingestion and merge-safe CLAUDE.md handling show thoughtful DX, but agent scaffolding is crowded.
OpenAPI sync prevents docs from rotting when Mintlify and GitBook require manual updates.
Tests agents on 700 policy docs and noisy voice calls where AgentBench stops.