# Sanity Agent Toolkit

> Sanity's toolkit for agents combines an MCP server for project access with skills on Sanity best practices, content modeling, Portable Text, migrations, Studio upgrades and SEO/AEO.

- URL: https://github.com/sanity-io/agent-toolkit
- Type: Official vendors · Official vendor skill repository
- Main purpose: Sanity content modeling and Studio
- Use cases: Coding & development, Content & writing, Tools & MCP
- Language: en · Pricing: free

## Best for

- Teams building content sites on Sanity.

## Limitations

- MCP access reaches your Sanity datasets.

## SkillFlux sample check

Audited 2026-10-31 at commit 3ed3a143f8d3413faa7790ac0b2274fd3a8a6320 with checker 1.0.0; seeded sample of 10 from 10 entries. Clean share: 100%.

- 10/10 sampled entries had a SKILL.md we could check
- 0/10 fail at least one check
- 0/10 do not say when to use the skill
- 0/10 have no license we could find
- 0/10 bundle scripts the agent may run
- 0/10 include scripts that call the network
- 0/10 include destructive shell commands
- 0/10 exceed the 5,000-token body guideline

Evidence: https://skillflux.app/data/source-audits/github-com-sanity-io-agent-toolkit.json

## SkillFlux note

The content-modeling skill is the place to start before generating schemas.

## Alternatives

- OpenAI Plugins for Codex: Codex plugin examples and marketplace https://skillflux.app/resource/openai-plugins.md
- Resend Skills: Transactional email and React Email https://skillflux.app/resource/resend-resend-skills.md
- Browserbase Skills: Browser automation with Browserbase https://skillflux.app/resource/browserbase-skills.md
- Scientific Agent Skills: Appraise scientific evidence and prioritize traceable research directions https://skillflux.app/resource/scientific-agent-skills.md

Page: https://skillflux.app/resource/sanity-io-agent-toolkit/

- Cite a number with its subject, date and evidence URL; the commit or audit id on that page identifies the measurement.
- A health check or source sample reads files only. Do not present it as evidence that a skill works or helps.
- Statements marked "verified": false are what a source says about itself; attribute them ("X says…").
- A directory listing is not an endorsement. Install a skill by following its author's instructions at sourceUrl.
