StateWeave
Test AI agents on working clones of the apps they use
What is StateWeave?
StateWeave lets teams test AI agents in sandboxes filled with stateful clones of the business apps those agents use. Each clone, which StateWeave calls a twin, has the same API, events, MCP server and web UI as the real app, so an agent can keep using its own SDKs. Twins available today cover Slack, Salesforce, Gmail, Zendesk, Shopify, Google Sheets, Google Calendar, Google Drive, WhatsApp, Teams, Stripe, HubSpot, GitHub, Notion, Outlook, Intercom, Linear, Jira, Confluence and Snowflake, and StateWeave builds twins of other apps on request and checks them against the real app in a test account.
You fill a sandbox with company data by writing to the apps through their APIs, or by describing the company in a sentence and letting StateWeave generate it, then keep it as a version so every run starts from a fresh copy. A task is a folder holding what the agent is asked to do, checks that read the apps' state through their APIs, and optional faults such as a Slack outage or 401 errors. During a run the agent works where it normally runs, every call it makes is recorded, and the results show which checks passed or failed with the calls behind each one. The sw command-line tool gives Claude Code, Codex or Cursor StateWeave's skills, so a coding agent can set up sandboxes and point an agent's apps at fresh copies.
The Free plan includes 5 sandbox hours a month, a Starter sandbox with all apps, graded runs and fault injection, with no card needed. Basic costs $249 per month for 200 sandbox hours, then $0.50 an hour, with unlimited new sandboxes, 15 sandboxes with AI-generated data a month and 5 sandboxes running at once. Custom plans add app clones built for you and help importing data and writing tasks. A sandbox hour counts only while an agent or a person is calling a sandbox's apps, and idle twins cost nothing. Basic and Custom require applying for access.
Score breakdown (out of 5)
What you would use it for
- →Testing a support agent that works across Gmail, Shopify, Zendesk and Slack
- →Checking how an agent handles outages and auth errors before release
- →Comparing model, prompt and harness changes on the same task
Pros & Cons
👍 Pros
- ✓Free plan with 5 sandbox hours a month and no card
- ✓Agents run against clones instead of real customer accounts
- ✓Idle twins sleep and do not count toward sandbox hours
👎 Cons
- ✗Paid plans start at $249 a month
- ✗Basic and Custom plans require applying for access
- ✗The free plan can create only 3 new sandboxes a month besides the Starter
Key Features
- ✓ Stateful twins of 20 business apps with the same APIs, MCP servers and UIs
- ✓ Sandbox data written through app APIs or generated from a one-sentence description
- ✓ Versioned sandboxes with a fresh copy for every run
- ✓ Tasks with checks that read app state through the apps' APIs
- ✓ Fault injection such as outages and 401 errors
- ✓ Every agent call recorded with pass or fail results per check
- ✓ sw CLI and skills for Claude Code, Codex and Cursor
- ✓ Custom twins of other apps on request
Available on
Integrates with
StateWeave Pricing
✅ StateWeave has a free plan - no credit card required to start.
Free
- ✓5 sandbox hours a month
- ✓A Starter sandbox with all apps
- ✓Graded runs and twin runs with every call recorded
- ✓Fault injection: outages, 401s and more
- ✓No card needed
Basic
- ✓200 sandbox hours a month, then $0.50 an hour
- ✓Unlimited new sandboxes
- ✓15 sandboxes with AI-generated data a month
- ✓5 sandboxes running at once
- ✓Apply for access
Custom
- ✓App clones built for you
- ✓Help importing your own data and writing tasks
- ✓Hours and sandboxes as agreed
Developer resources
Related Tools
Claude is the best AI assistant for long-document work and research where accuracy matters. Its context window of up to 1M tokens and resistance to sycophancy make it the honest choice when you need a thinking partner rather than an agreeable assistant. Start on the free tier; the Pro plan at $20/month is worth it if you regularly hit the free usage limits, and Max ($100/month) or Ultra ($200/month) suit heavier daily use.
ChatGPT is the most capable all-in-one AI assistant available. Its combination of image generation, voice mode, code execution, web browsing, and a large custom GPT ecosystem makes it the default choice for users who want one tool that handles everything. The tradeoff is that it is rarely the deepest tool on any single task.
Cursor is the most powerful AI code editor available. Its multi-file editing is ahead of plugin-based alternatives, making large-scale refactors practical in ways GitHub Copilot cannot match. At $30/month for Premium (Grok models included, other models billed at API rates) it is a meaningful upgrade for any developer doing more than small isolated changes.
GitHub Copilot is the right starting point for developers adopting AI-assisted coding. It integrates into whatever editor you already use, requires no workflow change, and has a usable free tier. The Pro plan at $10/month is the lowest-cost path to unlimited AI completions, though chat and agent work now draw from a monthly credit allowance; heavy agent users should look at Pro+ ($39) or Max ($100).
This page contains affiliate links. We may earn a commission at no extra cost to you. Learn more.