Taming 40+ MCP tools: Agent Skills for progressive discovery plus a single master execution tool
Longjumping_Bad_879 · reddit · 2026-10-01
A practitioner details a workaround for MCP context bloat when exposing 40+ tools.
The problem
- Deferred tool loading (expose few tools + a searchtool()) depends on host implementations, and keyword-based search becomes unreliable with many semantically overlapping tools—tool discovery is really a retrieval problem.
Agent Skills as progressive discovery
- Skills and reference files are structured as a multi-level index: intent → skill → reference → narrower reference → required tool. Routing is semantic and hierarchical; unneeded info never enters context.
Execution via a single master tool
- Skills handle discovery but not secure execution, so the author exposes only one facade tool (execute<businessapp>tool with internal tool name and args). Each internal tool's name, schema, usage guidance, examples, and constraints live in the Skills docs.
- Net effect: Skills own discovery/routing/instructions, MCP owns secure execution, and adding internal tools no longer grows the tool-definition context the model sees.
More from coding & agent
- AI Agent Polled Every 15 Minutes and Snagged a Fully-Booked 6-Seat Tokyo Restaurant in 13 Hours — armand_ruiz · 2026-10-01
- Higgsfield launches ChatGPT extension bringing computer use into Codex — SimplyAnnisa · 2026-10-01
- LibraryDesignBench tests whether AI agents can design and effectively use code libraries — a1zhang · 2026-10-01
- Ace launches as a digital employee with its own computer, email and logins — garrytan · 2026-10-01
- Anthropic's cost-optimize skill audits Claude API spend per completed task — udmrzn · 2026-10-01
- Dev rant: CI vs GitHub — the fierce fight over who slows you down — garrytan · 2026-10-01