llms.txt
Compare with other sites' llms.txt files →# Rystic
> Rystic builds local, stateful simulators of the third-party APIs your agents depend on — Stripe, GitHub, Kalshi — validated against the real service, and safe to reset, manipulate, and break.
Set up with Installation: https://www.rystic.ai/docs/installation.md. Agents then read https://www.rystic.ai/docs/agent-quick-start.md for what to report back, discovery and the machine-readable surfaces. https://www.rystic.ai/docs/what-you-can-do.md maps each task to the page that explains it.
Simulators available: Slack, GitHub, Kalshi, PayPal, Lob, Okta, Linear, Stripe, Resend, Anthropic, Clerk, Dropbox, Gmail, Google Drive, Jira, Notion, PostHog, QuickBooks, S3, Shopify, Word.
## Docs
- [Installation](https://www.rystic.ai/docs/installation.md): From nothing to a running simulator, for people and coding agents. Install, sign in, run, make a request, discover what a simulator serves.
- [What you can do](https://www.rystic.ai/docs/what-you-can-do.md): Common tasks with the simulators, each linked to the page that explains it.
- [Plans and API keys](https://www.rystic.ai/docs/licensing.md): Free, Pro or Enterprise — what each plan gets you, the 14-day Pro trial, how an API key signs the CLI in, what a running simulator sends, and what its errors mean.
- [Agent quick start](https://www.rystic.ai/docs/agent-quick-start.md): Written for the agent, not the human. What to do beyond Installation, what to report back, and the machine-readable surfaces.
- [Glossary](https://www.rystic.ai/docs/concepts/glossary.md): One sentence per term these docs use, with a link to the page that owns it.
- [Scenarios and seeding](https://www.rystic.ai/docs/concepts/worlds.md): Named scenarios, the shared design studio, seeding your own records with POST /_rystic/seed, the registers that hold them, rystic describe and the seed gate.
- [Reset and determinism](https://www.rystic.ai/docs/concepts/reset-and-determinism.md): Every boot and every reset starts from the same state. The pinned clock, the fixed RNG seed, pacing and what is and isn't deterministic.
- [Observing a simulator](https://www.rystic.ai/docs/concepts/observing.md): Read a simulator's state, the requests it served, a live event feed and what it would have sent out. Wait for your app to finish and browse it all in the state UI or the API explorer.
- [Snapshots and test cases](https://www.rystic.ai/docs/concepts/snapshots-and-test-cases.md): The baseline a run is measured from, snapshots to roll back to, scenario files a later run boots from and test cases that grade a later run, through rystic mcp or the CLI.
- [Examples and scripts](https://www.rystic.ai/docs/examples.md): The MIT-licensed scripts each simulator ships, with a run command for each — check.sh for every simulator, the Kalshi bots and checks and the GitHub review bots.
- [Multi-simulator environments](https://www.rystic.ai/docs/environments.md): Run several simulators as one project from a rystic.yaml — start, stop and reset them together, read and move their clocks together, follow them on one timeline, export the stack to Compose.
- [Intercept vendor calls](https://www.rystic.ai/docs/intercept.md): Route an unmodified app's calls for api.stripe.com, slack.com or s3.amazonaws.com to local simulators, with no code change and no base-URL override.
- [Webhooks and provider actions](https://www.rystic.ai/docs/webhooks.md): Test the outbound side of an integration across simulators — signed webhook delivery to your receiver, the egress log, provider actions and timelines.
- [Timelines](https://www.rystic.ai/docs/timelines.md): Schedule calls, seeds and messages on a simulator's clock, pass one call's response to the next with response bindings and read back what fired.
- [Network profiles and fault injection](https://www.rystic.ai/docs/network-profiles.md): Add latency and refuse a fraction of requests per endpoint with a network profile, from the control plane or at boot, and throttle Stripe charges with 429s.
- [Running a shared simulator](https://www.rystic.ai/docs/shared-simulators.md): One simulator for many clients or agents — lock the control plane with an admin token, log every change and expose it on a network.
- [Integration tests in CI](https://www.rystic.ai/docs/integration-testing.md): Run your integration tests against a simulator in GitHub Actions — one service container, one secret, your existing test runner.
- [Grade agent runs](https://www.rystic.ai/docs/agent-evals.md): Run your agent once per eval with rystic grid, each run on simulators of its own, grade every run PASS or FAIL with rystic grade and turn a good run into a regression eval.
- [Slack MCP in Harbor](https://www.rystic.ai/docs/slack-harbor.md): Connect your Harbor agent to Slack MCP. Get the licensed image, export a task, verify a reply and reset, then attach your own policy and grader, or run the reference harness.
- [Connect your coding agent](https://www.rystic.ai/docs/mcp-server.md): Connect Claude Code, Cursor, Codex, VS Code, Windsurf or Claude Desktop to rystic mcp, the local MCP server through which your coding agent starts, seeds, clocks and inspects simulators and keeps runs as test cases.
- [Rystic on Vercel](https://www.rystic.ai/docs/vercel.md): Connect Rystic once through Vercel Connect, pull a 24-hour API key from your code, and run a metered simulator in a Vercel Sandbox.
- [Slack](https://www.rystic.ai/docs/simulators/slack.md): A local Slack for bots and agents: conversations, messages, users, reactions, files, views, canvases, lists, search and Events API delivery.
- [AI-app threads](https://www.rystic.ai/docs/simulators/slack/features/ai-app-threads.md): Set an AI app thread's status, title and suggested prompts as Bolt's Assistant class does, and open the app's container as a person so your app hears assistant_thread_started.
- [Bot turns](https://www.rystic.ai/docs/simulators/slack/features/bot-turns.md): GET /_rystic/settle tells your driver when your bot's turn is over, and the call log shows what the bot did in it.
- [Events API](https://www.rystic.ai/docs/simulators/slack/features/events-api.md): Slack's Events API for your bot over Socket Mode or a request URL, with retries and app mentions.
- [Guests and Slack Connect](https://www.rystic.ai/docs/simulators/slack/features/guests-and-connect.md): Guests whose own tokens see only their channels and the people in them, and Slack Connect channels shared with members of another organization, as a bot reads them.
- [Identities](https://www.rystic.ai/docs/simulators/slack/features/identities.md): Several agents share one workspace, each under its own token, acting as a seeded user, with a call log for grading.
- [Interactivity](https://www.rystic.ai/docs/simulators/slack/features/interactivity.md): A person presses a button on your bot's message or submits its modal, and your app receives Slack's block_actions or view_submission payload.
- [MCP server](https://www.rystic.ai/docs/simulators/slack/features/mcp.md): The unmodified korotovsky Slack MCP server runs against the simulator through four tools: channel discovery, history, thread replies and posting a reply.
- [Search](https://www.rystic.ai/docs/simulators/slack/features/search.md): Search the messages a user can read with Slack's query modifiers, and files by text.
- [Verified UI](https://www.rystic.ai/docs/simulators/slack/features/verified-ui.md): A Slack client that runs on the simulator's Web API, run beside the simulator or as one container per task.
- [Verified UI tasks](https://www.rystic.ai/docs/simulators/slack/features/verified-ui-tasks.md): Seed a task's workspace and timed events, reset between episodes, grade what the agent did and size how many environments fit on one box.
- [What the verified UI supports](https://www.rystic.ai/docs/simulators/slack/features/verified-ui-support.md): Each part of the Slack client the verified UI draws, the Web API methods it calls and what it leaves out.
- [Workspace UI](https://www.rystic.ai/docs/simulators/slack/features/workspace-ui.md): A browser view of the workspace at /_rystic/ui that draws the simulator's state as a Slack client and redraws as your bot calls the API.
- [Slack changelog](https://www.rystic.ai/docs/simulators/slack/changelog.md): What changed in each Slack simulator release, newest first.
- [GitHub](https://www.rystic.ai/docs/simulators/github-pulls.md): A local GitHub for your client, git and gh: pull requests, reviews, review comments, merges, check runs, issues and a real git repository.
- [Auto-merge](https://www.rystic.ai/docs/simulators/github-pulls/features/auto-merge.md): Auto-merge requests that wait on a blocked pull request and land when its base's protection is met.
- [Branch protection and rulesets](https://www.rystic.ai/docs/simulators/github-pulls/features/branch-protection.md): Branch protection and rulesets that gate merges, and the mergeable state each pull request reads.
- [CI results and the evaluation engine](https://www.rystic.ai/docs/simulators/github-pulls/features/ci-evaluation.md): How each workflow job's result is decided without running it, from a recorded verdict or an evaluator you register, and the episodes an RL harness rewards on.
- [Commits through the API](https://www.rystic.ai/docs/simulators/github-pulls/features/commits.md): Branch and commit with no clone through the contents and git data APIs, in git's own shas, and how the REST API reads a commit's verification and authors.
- [Git repositories](https://www.rystic.ai/docs/simulators/github-pulls/features/git.md): A real git repository per scenario on the simulator's own port: clone, fetch, push and test merges, all with git's own shas.
- [GitHub Apps](https://www.rystic.ai/docs/simulators/github-pulls/features/github-apps.md): A GitHub App authenticating as itself: an app JWT, installation tokens and a bot user whose standing is the token's grant.
- [GraphQL and the gh CLI](https://www.rystic.ai/docs/simulators/github-pulls/features/graphql.md): The GraphQL subset the gh CLI speaks at POST /graphql, and the gh commands that run against the simulator.
- [Issue and pull request search](https://www.rystic.ai/docs/simulators/github-pulls/features/search.md): GitHub's issue and pull request search over REST and GraphQL, one qualifier grammar over what the scenario holds.
- [Issues, events and the timeline](https://www.rystic.ai/docs/simulators/github-pulls/features/issues-and-timeline.md): Issues and pull requests on one number sequence, their conversation comments, issue events, the merged timeline, rendered bodies and avatars.
- [Merge queues](https://www.rystic.ai/docs/simulators/github-pulls/features/merge-queue.md): Merge queues that build groups, run their checks and land them in order.
- [Protected pushes](https://www.rystic.ai/docs/simulators/github-pulls/features/protected-pushes.md): How branch protection and rulesets judge a git push, a ref write and a contents write, and the refusal text each one gets.
- [Pull request UI](https://www.rystic.ai/docs/simulators/github-pulls/features/pull-request-ui.md): The repository drawn as GitHub's own pull request pages in your browser, redrawn as your automation calls the API, with its own actions.
- [Review comment threads](https://www.rystic.ai/docs/simulators/github-pulls/features/review-comments.md): Line-anchored review comments on a pull request's diff, their replies and the refusal of a path the diff does not touch.
- [Review-bot scenarios](https://www.rystic.ai/docs/simulators/github-pulls/features/bot-scenarios.md): Two scenarios that hold a working market-maker bot as diffs and pull requests that each break it one way, for testing a review agent.
- [Workflows and CI runs](https://www.rystic.ai/docs/simulators/github-pulls/features/workflows.md): The repository's own workflows start on push and pull request, one check run per job, completed from an evaluation engine instead of running the job.
- [GitHub changelog](https://www.rystic.ai/docs/simulators/github-pulls/changelog.md): What changed in each GitHub simulator release, newest first.
- [Kalshi](https://www.rystic.ai/docs/simulators/kalshi.md): A local Kalshi exchange for trading bots: the Trade API v2, price-time-priority matching, the WebSocket feed, portfolio, P&L and tape replay.
- [Batch backtests](https://www.rystic.ai/docs/simulators/kalshi/features/batch-backtests.md): `rystic batch` replays every tape in a tape archive through its own simulator with your bot against each, natively or one container per tape, and ends in a per-market P&L table.
- [Browser UIs](https://www.rystic.ai/docs/simulators/kalshi/features/browser-uis.md): A state UI at /_rystic/ui for the book, P&L and order entry, plus Kalshi Accurate, a verified UI built against Kalshi's own interface.
- [Containers, Compose and offline runs](https://www.rystic.ai/docs/simulators/kalshi/features/containers.md): Runs the simulator as a container beside your bot, from a stack file or Compose and with the network off for an Enterprise session.
- [Exchange shards](https://www.rystic.ai/docs/simulators/kalshi/features/exchange-shards.md): Kalshi's numbered exchange shards: each market lives on one, an account holds cash per shard and an order debits its market's shard.
- [Live feed](https://www.rystic.ai/docs/simulators/kalshi/features/live-feed.md): Trades against a book that tracks the exchange right now: the CLI holds the exchange socket and feeds the simulator.
- [Live feed settings and Compose](https://www.rystic.ai/docs/simulators/kalshi/features/live-feed-settings.md): Every live-feed flag and its environment variable, Kalshi's other variables, and the live feed from a stack file or Compose, with a ticker or event list shared with your bot.
- [Order book and price grid](https://www.rystic.ai/docs/simulators/kalshi/features/order-book.md): How a book reads back, how to seed one and the per-market price grid that decides which prices an order may carry.
- [Orders and matching](https://www.rystic.ai/docs/simulators/kalshi/features/orders-and-matching.md): A price-time-priority matching engine behind Kalshi's V2 order routes, with retired V1 writes, fees, partial fills and two accounts trading one book.
- [Production latency](https://www.rystic.ai/docs/simulators/kalshi/features/production-latency.md): Makes the simulator respond as slowly as production Kalshi, using a per-endpoint latency profile measured against the live exchange.
- [Query errors](https://www.rystic.ai/docs/simulators/kalshi/features/query-errors.md): Reproduces the exchange's query-binding errors, in its field order, for the listed GET routes.
- [Sample bots, the grid and the bot contract](https://www.rystic.ai/docs/simulators/kalshi/features/bots-and-batches.md): Four sample bots, two smoke tests and an API check, `rystic grid`, which runs your bot across the simulator's scenarios, and the contract a bot keeps under the grid and `rystic batch`.
- [Scripted events](https://www.rystic.ai/docs/simulators/kalshi/features/scripted-events.md): Loads a schedule of messages, seeds and calls that fire on the simulator clock, such as a play-by-play feed or a settlement at the final out.
- [Settlement](https://www.rystic.ai/docs/simulators/kalshi/features/settlement.md): Resolves a market when you say so: cancels its resting orders, pays the winning side and writes the settlement history.
- [Several markets in one replay](https://www.rystic.ai/docs/simulators/kalshi/features/replay-groups.md): Replays correlated markets on one clock, games from different days back to back, or one tape as many markets for a load test.
- [Tape format, market records and replay limits](https://www.rystic.ai/docs/simulators/kalshi/features/tape-format.md): What a tape file holds, where a replayed market's record comes from, the one call to Kalshi that fetches it and what replay doesn't model.
- [Tape replay](https://www.rystic.ai/docs/simulators/kalshi/features/tape-replay.md): Plays a captured market day through the matching engine on the exchange clock, with your resting orders filled by the recorded flow.
- [Watching and grading a run](https://www.rystic.ai/docs/simulators/kalshi/features/grading-a-run.md): Prints state changes as they happen, logs every action to a file and gives a grader one-read account snapshots, all behind an optional admin token.
- [WebSocket feed](https://www.rystic.ai/docs/simulators/kalshi/features/websocket-feed.md): Serves Kalshi's WebSocket channels at the real path, reads back every frame the simulator rendered and cuts connections on demand.
- [Kalshi changelog](https://www.rystic.ai/docs/simulators/kalshi/changelog.md): What changed in each Kalshi simulator release, newest first.
- [PayPal](https://www.rystic.ai/docs/simulators/paypal.md): A local PayPal checkout for payment code: Orders v2, authorizations, captures, refunds and OAuth2 tokens, with seeded payment histories.
- [Payment lifecycle](https://www.rystic.ai/docs/simulators/paypal/features/payment-lifecycle.md): Orders, authorizations, captures and refunds keep PayPal's states, fee breakdowns and errors; an order is approved by a seed, and a card order can complete at creation.
- [PayPal Business UI](https://www.rystic.ai/docs/simulators/paypal/features/business-ui.md): The account drawn as PayPal's merchant dashboard in your browser, redrawn as your integration calls the API, with the merchant's own actions.
- [PayPal changelog](https://www.rystic.ai/docs/simulators/paypal/changelog.md): What changed in each PayPal simulator release, newest first.
- [Lob](https://www.rystic.ai/docs/simulators/lob.md): A local Lob for mail-sending code: addresses and postcards, idempotent sends and the cancellation window.
- [Dashboard](https://www.rystic.ai/docs/simulators/lob/features/dashboard.md): The mail account drawn as Lob's dashboard in your browser, redrawn as your code calls the API, with the dashboard's own actions.
- [Mail stream](https://www.rystic.ai/docs/simulators/lob/features/mail-stream.md): A postcard's journey after the send: USPS scans you record, and the cancellation window that closes the day after its send date.
- [Lob changelog](https://www.rystic.ai/docs/simulators/lob/changelog.md): What changed in each Lob simulator release, newest first.
- [Okta](https://www.rystic.ai/docs/simulators/okta.md): A local Okta org for provisioning code: users and their lifecycle, groups, application assignments, passwords and TOTP factors.
- [Bookmark applications](https://www.rystic.ai/docs/simulators/okta/features/applications.md): Bookmark applications are assigned to users directly or through groups, and a user's access follows its group membership and status.
- [Software TOTP factors](https://www.rystic.ai/docs/simulators/okta/features/factors.md): A user enrolls an OKTA software TOTP factor, activates it and verifies it with codes computed from the shared secret at the simulator's clock.
- [User lifecycle and passwords](https://www.rystic.ai/docs/simulators/okta/features/user-lifecycle.md): A user moves through Okta's statuses (staged, active, suspended, deprovisioned) by the lifecycle routes, and password routes expire, reset and change its password; an activation email is recorded, never sent.
- [Okta changelog](https://www.rystic.ai/docs/simulators/okta/changelog.md): What changed in each Okta simulator release, newest first.
- [Linear](https://www.rystic.ai/docs/simulators/linear.md): A local Linear workspace for agents: issues, comments, labels, relations, projects with milestones, documents and attachments over GraphQL.
- [Filters](https://www.rystic.ai/docs/simulators/linear/features/filters.md): The filter argument on issues, users and workflow states: comparators and connectives, plus one hop into related records on issues and workflow states.
- [GraphiQL explorer](https://www.rystic.ai/docs/simulators/linear/features/graphiql.md): GraphiQL in your browser over the schema the simulator serves, with each query run against the simulat