Fetching from the wire…
Public story · 2026-09-08 · high
The tool compiles MCP servers, OpenAPI specs, and databases into shell verbs a model loads one at a time.
Why now: declick's repo shows daily commits since it went up on September 2, 2026.
declick cut the tool-schema bytes an MCP client loads into context from 236,818 to 58,309 across nine servers with 258 tools, a 4.1x drop. Less schema in context up front means more room for the actual task, and fewer bytes to reprocess each turn.
declick compiles an OpenAPI spec, an MCP server, a database, or a web page into named shell verbs. It loads one verb at a time, not the full schema list up front. Each call returns a single response envelope with five exit codes.
declick ships with ten engines, no runtime dependencies, and runs on Node 24 under an Elastic-2.0 license. The repo includes node scripts/bench-tokens.mjs, so anyone running MCP servers can rerun it against their tool list and get an independent number.
The repo doesn't say how the shell-verb interface compares to a model's native tool-calling in practice. It also doesn't say whether byte savings map onto token savings, since bytes and tokens aren't the same unit.
Each link below shares sources, entities, or timing with this story.
Waishnav/devspace (4,247 stars, v1.0.8 August 25) runs a local Node server exposing read, edit, search and shell execution over MCP, reached through a reverse proxy tunnel like Cloudflare or ngrok, with password-gated owner approval per client. The pitch is turning ChatGPT int...
The 1.4.0a2 alpha ships langchain.mcp, an MCPAdapter turning any MCP server into tools you hand straight to create_agent. Connection handling is FastMCP's rather than a reimplementation, so a URL, a local stdio script, an in-process FastMCP server, a multi-server config, or a...
If you wrote an MCP server before July, it's on a protocol shape the maintainers have already removed. Not deprecated-with-a-migration-window. Removed from the spec. MCP lead maintainers David Soria Parra and Den Delimarsky published an updated roadmap on August 22, and the re...
lean-ctx is a Rust-based system that sits between AI coding tools and LLMs, compressing file reads by 60-99% and shell output by 60-95% using Tree-sitter AST parsing for 18 languages. Cached re-reads cost only 13 tokens. Works as a standard MCP server with 49 tools. Compatible...
Built by Tobias Lutke (Shopify co-founder). Combines BM25 full-text search, vector semantic search, and LLM re-ranking — all running locally via GGUF models. Includes a built-in MCP server for direct integration with AI agents. No cloud dependency. Install with npm install -g...
Anthropic invented a file convention. It's now shipping GA inside a competitor's product. Nobody wrote a spec, nobody held a standards meeting, it just happened. On July 29, GitHub made agent skills and MCP server support generally available in Copilot code review for all Pro,...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.