# agentsite-static-version=81ba0708-d105-46df-808f-e71b677c1118 # stateweave.net > AI-optimized mirror of stateweave.net containing 23 pages totalling 34,678 words of clean markdown content, structured data, and semantic HTML. Original source: https://stateweave.net. Last updated: 2026-11-17T13:00:28.067Z. Each page is available as HTML (with JSON-LD structured data) and Markdown (text-only, ideal for LLMs and RAG). ## Homepage - [StateWeave | Test AI agents on realistic copies of your apps](/content/site-root.html): Test AI agents on working twins of Slack, Gmail, Salesforce and more, filled with your data. Re-test every model, prompt and harness change. (664 words) ## Articles & Blog Posts - [docs/api/index.html](/content/docs/api/index.html) (5,617 words) - [Troubleshooting | StateWeave Docs](/content/docs/troubleshooting/index.html): Fixes for common StateWeave problems: connection and key errors, sandboxes in the wrong mode, tasks that cannot be read or proven, tests that need a re-check, run limits, empty answers, and clients that reach the real service. (2,240 words) - [Fill a sandbox | StateWeave Docs](/content/docs/fill/index.html): Put data into a StateWeave sandbox in edit mode: through the apps' own APIs and SDKs, as other people, with StateWeave's actions for what no API can do, or generated from a prompt while it is blank. (2,440 words) - [Run your agent | StateWeave Docs](/content/docs/run/index.html): Run your agent on fresh copies of a StateWeave sandbox: twin runs to develop against and judge with a task's checks, graded runs of uploaded tests with sw run, Python or HTTP, results, transcripts, reruns and CI. (2,897 words) - [Client libraries | StateWeave Docs](/content/docs/libraries/index.html): The setting each popular Slack, Google, Salesforce, Shopify, Zendesk, WhatsApp, Stripe, HubSpot and GitHub client library needs to work against a StateWeave run or sandbox, in Python and Node, as tested. (1,884 words) - [MCP servers | StateWeave Docs](/content/docs/mcp/index.html): Give an MCP-based agent, such as Claude Code, the apps of a StateWeave run as MCP servers: the apps that have one, the config file, a script for Claude Code and an MCP client example. (1,454 words) - [Sandboxes | StateWeave Docs](/content/docs/sandboxes/index.html): A StateWeave sandbox is a set of apps and their data. It starts blank, is filled in edit mode, and keeps a version in test mode that every run starts from. (1,900 words) - [sw command line | StateWeave Docs](/content/docs/cli/index.html): Reference for every sw command: login, sandbox, env, task, run, trace, worker, setup, and the low-level run commands, with their options, output and exit codes. (3,348 words) - [Terms of Service | StateWeave](/content/terms/index.html): The agreement between you and stateweave, Inc. for using StateWeave. (976 words) - [Build an agent | StateWeave Docs](/content/docs/agent/index.html): What the agent in a StateWeave run is, and three ways to get one running: adapt the agent you have, use Claude Code with no code, or start from a complete, tested Python agent. (2,063 words) - [Tasks | StateWeave Docs](/content/docs/tasks/index.html): Write a StateWeave task in Harbor's layout: instruction.md, task.toml, and checks through the apps' public APIs. Then judge your agent with it. (1,550 words) - [Quickstart | StateWeave Docs](/content/docs/quickstart/index.html): Install sw, make a blank StateWeave sandbox, fill it through Slack's API, write a task with checks, run your agent on fresh copies and judge it. (844 words) - [From your coding agent | StateWeave Docs](/content/docs/coding-agents/index.html): sw setup gives Claude Code, Codex or Cursor two StateWeave skills: your coding agent connects your agent to fresh copies of its apps, runs it, writes and uploads tasks, and finds why a run failed, without stopping to ask. (925 words) - [Upload a task | StateWeave Docs](/content/docs/upload/index.html): Upload a task to your StateWeave workspace as a test: prove it with sw task import --check, keep it, and re-check it when the sandbox gets a new version. Uploading is the only way to add a test. (747 words) - [Faults | StateWeave Docs](/content/docs/faults/index.html): The failures real APIs have that a StateWeave task can inject, outages, 401s for valid tokens, hung calls, writes that commit but answer an error: the kinds, the presets, how a task's task.toml names them, how they are proven and how a run shows them. (905 words) - [Core concepts | StateWeave Docs](/content/docs/concepts/index.html): StateWeave's building blocks: workspace, apps (twins), sandboxes with edit and test modes and versions, tasks and their checks, twin runs and test runs, persona, faults and agent labels. (1,128 words) - [Privacy Policy | StateWeave](/content/privacy/index.html): What personal data StateWeave keeps, what it is used for, and your choices. (688 words) - [Security | StateWeave](/content/security/index.html): What StateWeave holds, where it is kept, who can reach it, and how to report a problem. (474 words) - [StateWeave documentation | StateWeave Docs](/content/docs/index.html): StateWeave gives your AI agent working copies of Slack, Gmail, Google Workspace, Salesforce, Shopify, Zendesk, WhatsApp, Microsoft Teams, Stripe, HubSpot and GitHub. Fill a sandbox, write tasks judged through the apps' APIs, and run your agent on fresh copies. (702 words) - [Pricing | StateWeave](/content/pricing/index.html): StateWeave pricing: Free for everyone with 5 sandbox hours a month, Basic at $249 a month, and Custom plans with app clones built for you. Apply for access to Basic and Custom. (324 words) - [Refunds and Cancellation | StateWeave](/content/refunds/index.html): How StateWeave's plans are billed, how to cancel, and when a charge is refunded. (320 words) ## About Pages - [Work with teammates | StateWeave Docs](/content/docs/team/index.html): Share a StateWeave workspace: send its invite link, let people in, decide who may do what, remove people, and switch between workspaces. (588 words) ## Resources - [Full Page Index](/index.html): Browse all cached pages with rich metadata - [About This Cache](/about.html): Methodology, technical details, and usage guidelines - [XML Sitemap](/sitemap.xml): Machine-readable sitemap for crawler discovery - [Robots.txt](/robots.txt): Crawler directives - [Agentsite Network](https://agentsite.network/network.html): Public index of Agentsites and their machine-readable resources