KnowledgeMCP – Turn any docs into an MCP endpoint (0 LLM at query time)
Zero LLM cost at query time when most MCP doc servers burn tokens on every search.

130 MCP tools for ATS management when competitors ship 30-40.
Recruiting teams using AI assistants like Claude or Cursor
GitHub MCP Server · Notion MCP · Zapier AI Actions
I would love HN to weigh in on whether 130 tools are too many. Big competitors have 30-40
Some details:
Three layers, from the bottom up: - REST API v2: 87 endpoints, OpenAPI 3.0 spec, Bearer auth, public docs at 100hires.com/api - MCP server: 130 tools mapping roughly 1:1 to REST endpoints, hosted at mcp.100hires.com/mcp, ~400 LOC of wrapper code - ChatGPT plugin: thin wrapper on the MCP, OAuth handled by the plugin store
Auth on the MCP: OAuth 2.1 with PKCE and Dynamic Client Registration. Tokens scoped per user, 1-hour access tokens with auto-refresh, revoked when the user leaves the workspace. Tested in Claude Desktop, Claude Code, Cursor, VS Code, Codex, Windsurf, and Zed.
Available on the free trial (but rate-limited) and on all paid plans.
Zero LLM cost at query time when most MCP doc servers burn tokens on every search.
Model-neutral MCP server with Docker sandbox and SWE-bench evals for any coding agent.
Sandboxed MCP server lets LLMs run Ghidra and Radare2 without blowing up your host.
rr reverse debugging as an MCP server—smart MCP integration, Linux-only limits reach.
VLCs over VLAs: LLMs write Python code against live robots instead of predicting actions.
Personal profile MCP with scrapers, but the connectors are limited and LLM agent use cases are early.