# Toolwitness

> Snapshot exercise evidence for AI-agent tools. 0 fictional examples; 0 reproduced results. 1 offline listings (excluded from exercised results).

Records are at /tool/<slug>/. Full category lists are at /category/<id>/; types at /type/<hooks|skills|mcp>/. Review metadata and raw artifacts accompany each record.

## Categories

- [developer-tools](https://toolwitness.ai/category/developer-tools/): 1 records.

## Agents

- [Claude Code](https://toolwitness.ai/agent/claude-code/): 0 records.
- [Codex](https://toolwitness.ai/agent/codex/): 0 records.
- [Gemini CLI](https://toolwitness.ai/agent/gemini-cli/): 0 records.
- [Grok Build](https://toolwitness.ai/agent/grok-cli/): 0 records.

## Main sections

- [Home](https://toolwitness.ai/): Bounded recent snapshot sample.
- [All snapshots](https://toolwitness.ai/snapshots/): Complete linked archive.
- [Found](https://toolwitness.ai/found/): Discovery only; not yet witnessed.
- [About](https://toolwitness.ai/about/): Admission and evidence states.
- [Method](https://toolwitness.ai/method/): Operational steps, scope and disclaimer.
- [hooks](https://toolwitness.ai/type/hooks/): Local hooks records.
- [skills](https://toolwitness.ai/type/skills/): Local skills records.
- [mcp](https://toolwitness.ai/type/mcp/): Local mcp records.

This text index is a convenience, with no ranking or citation promise.

## Recent machine-checked records

- [ai.aquex/reasonbook](https://toolwitness.ai/tool/offline-mcp-baca48c49bc4ef8ad1008cb8/): offline_listing; 2026-10-16T05:18:47.546326Z. Limits: Initialize and tools/list only. No tool called. No model ran. No network-attempt tracing or credential test. Later versions require a new run.