# Tools — Agent Field Notes

Agentic tools worth knowing about: what each does, what you gain from it, and the sources behind it.

## Anubis

- Website: https://github.com/uncsoft/anubis-oss/
- Code: https://github.com/uncsoft/anubis-oss/
- Category: Local models
- Added: 2026-10-25

Mac app for testing and benchmarking local models on Apple Silicon, with live power, thermal and memory readings during inference.

**What you gain:** Shows whether a long agent run will throttle your machine, not just how fast a single answer is. Results can be exported or posted to a community leaderboard.

Sources: [anubis-oss on GitHub](https://github.com/uncsoft/anubis-oss/)

## asiai

- Website: https://pypi.org/project/asiai/
- Category: Local models
- Added: 2026-10-25

Command-line tool that compares local AI inference engines side by side on an Apple Silicon Mac, with community benchmarks.

**What you gain:** Find the fastest engine for your own Mac before you build a workflow on it, and compare your numbers with other users on the same chip.

Sources: [asiai on PyPI](https://pypi.org/project/asiai/)

## Is it agent ready? (Cloudflare)

- Website: https://isitagentready.com
- Category: Agent readiness
- Added: 2026-10-25

Free scanner from Cloudflare that scores any website on how ready it is for AI agents, with fix-it prompts you can hand to a coding agent.

**What you gain:** If customers start sending AI agents to buy or book for them, a low score means lost sales. It takes a minute to check your own site.

Sources: [Softprom: Cloudflare Agent Readiness Score](https://softprom.com/cloudflare-agent-readiness-score-is-your-site-ai-ready-in-2026)

## MCP Doctor

- Website: https://glama.ai/mcp/servers/Jiansen/mcp-doctor/tree
- Category: MCP
- Added: 2026-10-25

Checks the quality of an MCP server — tool naming, descriptions, metadata for directories — and suggests concrete fixes.

**What you gain:** If you publish an MCP server, this tells you what stops directories and agents from picking it up, before you launch.

Sources: [MCP Doctor on Glama](https://glama.ai/mcp/servers/Jiansen/mcp-doctor/tree)

## MetriLLM

- Website: https://metrillm.dev
- Category: Local models
- Added: 2026-10-25

One-command benchmark for local models that scores speed, quality and hardware fit, with a shareable result and public leaderboard.

**What you gain:** A quick way to check whether a model is usable on your hardware before committing an agent job to it. It also runs as an MCP server, so an agent can benchmark for you.

Sources: [MetriLLM on Glama](https://glama.ai/mcp/servers/nyqwwnlvse)