local-first capability discovery

Stop loading
every tool.

Index skills, prompts, agents, and MCP servers already on your machine. Ask for a task. Load only the few resources that fit.

Try the search lab ↓View source ↗
zero-setup demo
$ npx --yes github:mockingbird777/task2tool demo

1. Pull Request Risk Reviewer skill · 14.273
  Reviews diffs for security regressions…

2. github mcp-server · 5.403
  Inspect repositories and pull requests…

2 matches from 8 example resources · local only
Interactive sample

Search a capability shelf.

This browser demo uses a small illustrative catalog. The CLI performs the same style of ranking against files on your machine.

3 best matches · 8 resources indexedBM25-inspired · deterministic
8resources indexed locally
3 / 8resources selected for context
Small surface, useful defaults

Built for the agent layer.

01

Scan what you have

Discover SKILL.md, agent and prompt files, MCP configurations, and portable JSON catalogs recursively.

02

Retrieve what matters

Weighted fields and BM25-style scoring turn natural-language tasks into a short, explainable result list.

03

Share without a service

Export deterministic JSON, readable Markdown, or a self-contained searchable HTML report. Zero runtime dependencies.

Your context window is a workspace, not a warehouse.
No cloud account, hosted index, telemetry, or API key
MCP arguments and environment values stay out of reports
Symlinks and generated dependency directories are skipped
Strict TypeScript and Node 20 with no runtime packages