Toollens — Development skill for Claude Code
Give your agent 2,468 tools.
How to install Toollens
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open krishnashakula/toollens and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Toollens does
Give your agent 2,468 tools. It sees 3. A Rust MCP gateway: dependency-aware tool discovery, policy-checked execution, measured on 110 real vendor MCP servers (26/28 tasks at half the cost of Anthropic's tool search).
Alternatives in Development
- Bypass 403 — Probe a 403/401 endpoint with the most-paid bypass tricks (header injection, path encoding, method swap, WAF f 4.5k ★
- Claudes C Compiler — Claude Opus 4.6 wrote a dependency-free C compiler in Rust, with backends targeting x86 (64- and 32-bit), ARM 2.6k ★
- Gov Backlinks — Free .gov backlinks - drive the user's browser through SAM.gov, the SBA small business listing, and state vend 195 ★
README
🔍 ToolLens
Give your agent 2,468 tools. It sees 3.
**An MCP gateway in Rust. Put any number of MCP servers behind it: your agent discovers, inspects and calls exactly the tools it needs, and finding a tool never grants permission to run it.**
[](https://github.com/krishnashakula/toollens/actions/workflows/ci.yml) [](LICENSE) [](https://www.rust-lang.org/) [](https://modelcontextprotocol.io/) [](CONTRIBUTING.md)
Real run, not a mock-up: Claude Opus 5.5 against 110 vendor-published MCP servers. Full 4K/60 fps video in the v0.1.0 release.
MCP clients send **every tool definition with every request**. Connect enough servers and that stops working: the 110 real vendor servers in this repo's benchmark expose 2,453 tool definitions, which come to **1.43M tokens per request, and the API rejects every call** before the model runs. Below that limit you still pay for, and distract the model with, hundreds of tools it will never use.
ToolLens sits between the agent and your servers and shows the agent three tools instead:
| Tool | What it does |
|---|---|
discover_tools(goal, limit?) |
Up to 10 tool ids for a goal: direct matches plus the tools they depend on, each with the reason it is there |
inspect_tool(tool_id) |
Full input/output schema, server provenance, and the current policy decision |
| `i |
Related Skills
Nemoclaw Runtime
A secure, always-on runtime for autonomous agents — gateway, policy guardrails, OpenShell sandbox, and durable
Meta Ads Policy Review
Pre-flight Meta ad policy checker for Claude — Cleared / Fix / Grounded verdicts with live policy citations an
Telltale
Dispatch room for five vendor CLIs. One brief, answered side by side. Statusline and HUD; every number is meas
Agentjail
Policy guardrails for coding agents (Claude Code, Codex, Cursor) — every tool call is checked locally, before
Voipi
🎙️ Give your apps, CLIs, and agents a voice. VoiPi is a universal, zero-dependency, free text-to-speech libra
Telepath
Give your Claude Code and Codex sessions telepathy — a zero-dependency message bus that lets local coding agen
Related Agents
Eval Inline Fact
Eval arm inline-fact (Q4 — the CALCULATED vs MEASURED line reduced to its FACT half: what the quantities are,
API Debugger
Use this agent for API gateway debugging — HTTP errors, auth, controller/routing issues, mock servers, caching
Close Reviewer
Independent reviewer for a prepared accounting close. Spawned by the close skill for the review phase, or dire