{"id":"030b002a-7388-47ad-928a-6bd7432b27b7","slug":"stuntdouble-mcp","kind":"mcp","name":"stuntdouble-mcp","summary":"MCP Service for Stunt Double","owner":"stunt-double","repoUrl":"https://github.com/stunt-double/stuntdouble-mcp","license":"MIT","licenseTier":"permissive","trust":"community","stars":1,"match":null,"fewStars":true,"format":"mcp","createdAt":"2026-03-26T14:40:49.000Z","addedAt":"2026-10-14T06:11:10.980Z","logoUrl":"https://avatars.githubusercontent.com/u/187224034?v=4","linkOnly":false,"collection":null,"skillPath":null,"categories":["developer-tools","design-media"],"tags":["ai","ai-agents","claude","claude-code","cursor","ux-design"],"runtime":"remote","local":false,"cloudCapable":false,"cloudCapability":"unknown","cloudReadyOnPlatform":false,"needsApiKey":false,"reviewed":false,"syncedAt":"2026-10-14T19:43:29.854Z","sample":false,"url":"https://conduits.dev/tools/stuntdouble-mcp","language":"JavaScript","transports":["http"],"install":{"remoteUrl":"https://app.stuntdouble.io/api/mcp"},"requirements":{"env":[],"upstreamMayCharge":false},"tools":null,"readme":"

\n \"Stunt\n

\n\n# Stunt Double MCP Server\n\nDeploy AI user personas to validate user journeys at scale. Find UX friction before real users do.\n\n[Stunt Double](https://stuntdouble.io) deploys AI agents with realistic user personas to validate user journeys at scale. Create actors, run automated workflows and checklists against any web app, and surface friction points before real users encounter them. Integrates with Claude, Linear, GitHub, and Slack.\n\n

\n \"Asking\n

\n\n## This repository\n\nThis repo holds **plugin and MCP configuration** (`.claude-plugin/plugin.json` and `.mcp.json` for Claude, `.cursor-plugin/plugin.json`, `marketplace.json` and root `mcp.json` for Cursor, `server.json` for the MCP registry), plus skills, agents, and Cursor rules. There is **no `package.json`** and **no runnable server** here: the MCP endpoint is hosted at `https://app.stuntdouble.io/api/mcp` from the main Stunt Double codebase. See [CONTRIBUTING.md](./CONTRIBUTING.md) for how to validate edits and avoid confusing this folder with a Node package.\n\n## Quick Start\n\n### Claude Code\n\n```bash\nclaude mcp add --transport http stuntdouble https://app.stuntdouble.io/api/mcp\n```\n\n### Claude Code plugin\n\nThe repository is also a Claude plugin: it bundles the hosted MCP server (`.mcp.json`) with the skills in `skills/` and the agents in `agents/`.\n\n```bash\nclaude plugin marketplace add stunt-double/stuntdouble-mcp\nclaude plugin install stuntdouble@stuntdouble\n```\n\n### Skills (any agent)\n\nThe skills in `skills/` install into Claude Code, Cursor, Codex, OpenCode and other agents that read `SKILL.md` files, using the [skills CLI](https://skills.sh). They drive the MCP server above, so connect that too.\n\n```bash\nnpx skills add stunt-double/stuntdouble-mcp # pick from the list\nnpx skills add stunt-double/stuntdouble-mcp --skill verify-change # just one\n```\n\n| Skill | What it does |\n| ----------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |\n| `check-agent-readiness` | Check how well AI agents (ChatGPT, Claude, Gemini, Perplexity and others) can find, understand and act on a website using the Stunt Double Index, explain the score from real agent sessions, compare against peers, and re-score after fixes |\n| `check-brand` | Audit a product or site against brand and tone-of-voice guidelines |\n| `check-compliance` | Check a product against legal and compliance requirements on Stunt Double (cookie consent, privacy and terms access, required disclosures, claim substantiation, unsubscribe flows) and collect evidence for counsel to review |\n| `check-continuity` | Check continuity across surfaces on Stunt Double |\n| `check-design-system` | Audit a live product against its design system on Stunt Double |\n| `create-actor-panel` | Create and configure a Stunt Double actor (AI persona) with knowledge entries for realistic user simulation |\n| `design-review` | Run a design review session by gathering feedback from multiple Stunt Double actors on a proposed design or flow |\n| `maintain-automations` | Change an existing Stunt Double automation or checklist without losing run history: retime triggers, edit and rewire workflow steps, build condition branches, update checks in place, and pause or retire what is no longer needed |\n| `run-qa-suite` | Run the full Stunt Double QA suite |\n| `run-user-interview` | Plan, configure, and launch a structured user interview with AI participants on Stunt Double, then read back the synthesised report |\n| `run-ux-validation` | Validate a user journey by running Stunt Double workflows or checklists and reporting the results |\n| `setup-guardrails` | Stand up continuous guardrails on Stunt Double |\n| `triage-feedback` | Review, categorize, and manage Stunt Double feedback submissions across projects |\n| `verify-change` | Verify a shipped or previewed code change by running a Stunt Double actor through the affected user flows, and optionally report results on the pull request |\n\n### Claude (web, Desktop, mobile)\n\nGo to **Settings → Connectors → Add custom connector** and paste:\n\n```\nhttps://app.stuntdouble.io/api/mcp\n```\n\n### ChatGPT and Codex\n\nThe repository also carries an OpenAI plugin manifest (`.openai-plugin/plugin.json`, in the [Agent Plugins](https://agent-plugins.org) format) with the directory listing, review test cases and release notes. `node scripts/build-openai-plugin.mjs` checks it against OpenAI's limits and writes `dist/stuntdouble-openai-plugin.zip` (in the Codex layout the portal accepts: the manifest as `.codex-plugin/plugin.json`, a `.mcp.json` for the hosted server, the skills and the logos), which is what goes into the [plugin submission portal](https://developers.openai.com/plugins/deploy/submission).\n\nUntil it is published in the directory, connect it in ChatGPT developer mode (**Settings → Apps → Advanced**) as a custom connector with:\n\n```\nhttps://app.stuntdouble.io/api/mcp\n```\n\nThe checklist run and Index report tools render as cards in ChatGPT (and in other MCP Apps hosts, Claude among them).\n\n### Codex CLI\n\n```bash\ncodex mcp add stuntdouble --url https://app.stuntdouble.io/api/mcp\ncodex mcp login stuntdouble\n```\n\nOr add it to `~/.codex/config.toml` by hand and then run `codex mcp login stuntdouble`:\n\n```toml\n[mcp_servers.stuntdouble]\nurl = \"https://app.stuntdouble.io/api/mcp\"\n```\n\n### Gemini CLI\n\n```bash\ngemini mcp add --transport http stuntdouble https://app.stuntdouble.io/api/mcp\n```\n\nOr add it to `~/.gemini/settings.json` (or `.gemini/settings.json` in a project):\n\n```json\n{\n \"mcpServers\": {\n \"stuntdouble\": {\n \"httpUrl\": \"https://app.stuntdouble.io/api/mcp\"\n }\n }\n}\n```\n\nRun `/mcp auth stuntdouble` inside Gemini CLI to sign in.\n\n### VS Code (GitHub Copilot)\n\n```bash\ncode --add-mcp '{\"name\":\"stuntdouble\",\"type\":\"http\",\"url\":\"https://app.stuntdouble.io/api/mcp\"}'\n```\n\nOr add it to `.vscode/mcp.json` in a workspace (or run **MCP: Open User Configuration** for every workspace):\n\n```json\n{\n \"servers\": {\n \"stuntdouble\": {\n \"type\": \"http\",\n \"url\": \"https://app.stuntdouble.io/api/mcp\"\n }\n }\n}\n```\n\nThe tools appear in Copilot Chat's agent mode.\n\n### OpenCode\n\nAdd it to `opencode.json` in a project (or `~/.config/opencode/opencode.json`), then run `opencode mcp auth stuntdouble`:\n\n```json\n{\n \"$schema\": \"https://opencode.ai/config.json\",\n \"mcp\": {\n \"stuntdouble\": {\n \"type\": \"remote\",\n \"url\": \"https://app.stuntdouble.io/api/mcp\"\n }\n }\n}\n```\n\n### Windsurf\n\nAdd it to `~/.codeium/windsurf/mcp_config.json` (or **Windsurf Settings → Cascade → MCP Servers → View raw config**):\n\n```json\n{\n \"mcpServers\": {\n \"stuntdouble\": {\n \"serverUrl\": \"https://app.stuntdouble.io/api/mcp\"\n }\n }\n}\n```\n\n### Other clients\n\nAny client that speaks Streamable HTTP and OAuth can connect to `https://app.stuntdouble.io/api/mcp` directly. For a client that only runs local (stdio) servers, bridge it with [`mcp-remote`](https://www.npmjs.com/package/mcp-remote), which handles the OAuth sign-in:\n\n```json\n{\n \"mcpServers\": {\n \"stuntdouble\": {\n \"command\": \"npx\",\n \"args\": [\"-y\", \"mcp-remote\", \"https://app.stuntdouble.io/api/mcp\"]\n }\n }\n}\n```\n\n### Cursor\n\nThis repository is a [Cursor plugin](https://cursor.com/docs/reference/plugins.md): `.cursor-plugin/plugin.json` and `.cursor-plugin/marketplace.json`, root `mcp.json` (the hosted MCP server), plus the skills in `skills/`, rules in `rules/`, and agents in `agents/`. The plugin `logo` is `assets/logo.png`. Auth is OAuth 2.1 with PKCE; no API keys.\n\n**Install as a plugin (recommended):** once published, install **Stunt Double** from the [Cursor marketplace](https://cursor.com/marketplace). For a team marketplace or local clone, add this repo as a marketplace and install the `stuntdouble` plugin; Cursor discovers the MCP server, skills, rules and agents from the default folders.\n\n**MCP only (no skills, rules or agents):** add the block below to a Cursor MCP config:\n\n- **Project-local:** `.cursor/mcp.json` at the root of your project.\n- **Global (all projects):** `~/.cursor/mcp.json` on macOS/Linux (see [Cursor MCP docs](https://cursor.com/docs/mcp.md) for your OS).\n\n```json\n{\n \"mcpServers\": {\n \"stuntdouble\": {\n \"url\": \"https://app.stuntdouble.io/api/mcp\"\n }\n }\n}\n```\n\nUse only `url` for remote servers (Streamable HTTP is negotiated automatically). Extra keys such as `\"type\": \"streamable-http\"` are not part of [Cursor's documented `mcp.json` shape](https://cursor.com/docs/mcp.md) and can break plugin validation.\n\nThe `server.json` file is the separate [MCP registry](https://modelcontextprotocol.io/registry/about) manifest for `mcp-publisher` and directory listings; Cursor's installer does not use it.\n\n## Authentication\n\nAuthentication is handled automatically via OAuth 2.1 with PKCE. The first time your AI client connects, a browser window will open for you to sign in and authorise access to your Stunt Double account. No API keys or tokens required.\n\nFor **Cursor**, the OAuth redirect URI is fixed to `cursor://anysphere.cursor-mcp/oauth/callback` ([docs](https://cursor.com/docs/mcp.md)).\n\n### What a connection can reach\n\nThe server acts as you, never more. Every tool resolves the workspace it is\nbeing asked about and checks your membership of it before doing anything, so a\nconnection reaches exactly the workspaces `list_workspaces` returns for you, and\nan archived workspace reaches nothing. A workspace you are not a member of\nanswers the same way one that does not exist does: \"not found\".\n\nThat holds for ids too, not just the `workspace_id` you pass. A tool that takes\nanother object's id (a checklist for an automation step, an actor for an\ninterview participant) checks that object belongs to the same workspace before\nstoring or running it, and refuses with \"not found in this workspace\" otherwise.\nRefusals never say which workspace an id does belong to.\n\nScopes narrow this further, never widen it. The consent screen names what the\nconnection asked for, and a token granted `mcp:read` is not shown the write or\nrun tools at all: they are absent from `tools/list` rather than present and\nfailing.\n\n| Scope | What it allows |\n| ----------- | --------------------------------------------------------------------------------------------------------------------------------------------------------- |\n| `mcp:read` | Read your workspaces and their contents |\n| `mcp:write` | Create and edit content in your workspaces |\n| `mcp:run` | Start checklist runs, automation runs and interviews, which consume the workspace run allowance, and re-run Stunt Double Index scores for domains you own |\n\n## Available Tools\n\n### Account\n\n| Tool | Description |\n| -------- | ------------------------------------------------------------------------------------ |\n| `get_me` | The account this connection acts as: id, email, name, timezone, notification channel |\n\nNotification channel and timezone are account settings rather than workspace ones, so\nthey are the same answer in every workspace. Pass the `timezone` when creating a\nscheduled workflow, or \"every weekday at 9\" becomes nine in UTC.\n\n### Workspaces\n\n| Tool | Description |\n| ------------------------ | ---------------------------------------------------------------------------------- |\n| `list_workspaces` | List your workspaces |\n| `get_workspace` | Get workspace details by ID or slug, including its admin-set controls (`settings`) |\n| `list_workspace_members` | List members of a workspace |\n\n`get_workspace` reports the workspace security controls under `settings`: public\nsharing, the feedback widget, self-hosted workers, and the network policy. They are\nceilings set by an admin, so a feature switched off there cannot be switched back on for\na single project.\n\n### Search\n\n| Tool | Description |\n| -------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |\n| `search` | Search a workspace across projects, actors, checklists, interviews, automations, issues, goals, feedback, actor knowledge, insights, design reviews, project resources and conversations. Ranked, with optional filters |\n\n`search` takes a `workspace_id` plus an optional `query`, `types` filter, `project_id`\nfilter and `limit` (max 50). Terms are matched as prefixes, so a partial word is enough.\nOmit `query` to browse the most recently updated items. Results carry the `id` you need\nfor the matching getter, so it is usually cheaper than listing an entity type and\nfiltering the list yourself. Reach for it before creating anything, to find the actor or\nchecklist that already covers the job.\n\n### Projects\n\n| Tool | Description |\n| -------------------------- | -------------------------------------------------------------------------- |\n| `list_projects` | List projects in a workspace |\n| `get_project` | Get a project (the product tracked by checklists, workflows, feedback) |\n| `create_project` | Create a project (a product to track with checklists, workflows, feedback) |\n| `list_project_mcp_servers` | The MCP servers this project's runs can reach |\n\nA project is archived, never deleted, and an archived project reads as missing from\nevery tool here. Registering an MCP server and attaching it to a project are\nworkspace-admin actions in the dashboard; `list_project_mcp_servers` is how you check\nwhat tools a run will actually have before writing a checklist that depends on one.\n\n### Guidelines\n\nStanding rules the team holds the product to: design system, tone of voice, brand,\ncontent, accessibility, compliance, security, performance, or shared knowledge. A\nguideline is owned by the **workspace** and attached to the projects it applies to, so\none rule can hold for every project without being retyped. Whatever is in force is\nrendered into every checklist run, design review, interview and triage for that project.\n\n| Tool | Description |\n| ---------------------------- | ------------------------------------------------------------------------------- |\n| `list_workspace_guidelines` | The workspace library, with how many projects hold each rule |\n| `add_workspace_guideline` | Add a rule to the library, optionally attaching it to projects |\n| `update_workspace_guideline` | Edit a rule, switch it off, or apply it to every design review in the workspace |\n| `remove_workspace_guideline` | Remove a rule from the library, detaching it from every project |\n| `list_project_guidelines` | The rules this project is held to |\n| `add_project_guideline` | Record a rule and hold this project to it |\n| `set_project_guideline` | Attach a library rule to a project, detach it, or switch it off there |\n\nTwo switches decide whether a rule is in force for a project: the library's `enabled`\nand the attachment's. `list_project_guidelines` folds them into one `enabled` so you\nnever have to reason about both. A rule with `apply_to_design_reviews` set also holds for\ndesign reviews raised from Slack or Linear, which carry no project to attach it through.\n\nCodify a standard as a guideline rather than repeating it in each checklist, and search\nthe library before writing a new rule: attaching the one that already exists keeps the\nteam's standard in a single place to edit.\n\n### Actors\n\n| Tool | Description |\n| -------------- | --------------------------------------------------------------------------------------------------------------- |\n| `list_actors` | List active actors in a workspace |\n| `get_actor` | Get actor details including system prompt and capabilities |\n| `create_actor` | Create a new actor in a workspace |\n| `update_actor` | Update actor name, description, system prompt, capabilities, or status. Set status to \"archived\" to soft-delete |\n\n### Knowledge\n\n| Tool | Description |\n| ------------------------ | ----------------------------------- |\n| `list_actor_knowledge` | List knowledge entries for an actor |\n| `add_actor_knowledge` | Add a knowledge entry to an actor |\n| `remove_actor_knowledge` | Remove a knowledge entry |\n\n### Conversations\n\n| Tool | Description |\n| -------------------- | ------------------------------------------------ |\n| `list_conversations` | List conversations, optionally filtered by actor |\n| `get_conversation` | Get a conversation with its messages |\n\n### Checklists\n\n| Tool | Description |\n| ------------------- | ------------------------------------------------------------------------------------------ |\n| `list_checklists` | List checklists in a workspace |\n| `get_checklist` | Get checklist details, checks, and recent runs |\n| `get_checklist_run` | Get a checklist run with per-check results |\n| `run_checklist` | Trigger a checklist run (async). Returns run ID |\n| `create_checklist` | Create a browser-based QA checklist (host via project or URL, actor, instructions, checks) |\n| `update_checklist` | Update a checklist (pass `checks` to replace the full set) |\n| `delete_checklist` | Delete a checklist and its checks and runs |\n\n### Workflows\n\nA workflow is a graph, not a list: its steps run by following the connections between them. The step tools maintain those connections, so adding, removing and reordering steps is enough to build one. `connect_workflow_steps` is only needed to branch.\n\n| Tool | Description |\n| ------------------------ | ---------------------------------------------------------------------------- |\n| `list_workflows` | List workflows in a workspace |\n| `get_workflow` | Get a workflow with its steps, edges, the order a run takes, and recent runs |\n| `run_workflow` | Trigger a workflow run (async). Returns run ID |\n| `get_workflow_run` | Get a workflow run with step-level details |\n| `create_workflow` | Create a workflow (multi-step automation) |\n| `update_workflow` | Update a workflow's name, description, or trigger |\n| `toggle_workflow` | Activate or pause a workflow |\n| `delete_workflow` | Delete a workflow and its steps and runs |\n| `add_workflow_step` | Add a step and connect it into the run |\n| `update_workflow_step` | Change a step's type or config in place |\n| `remove_workflow_step` | Remove a step and close the gap it leaves |\n| `reorder_workflow_steps` | Set the order the steps run in |\n| `connect_workflow_steps` | Wire one step to another, for a condition's True and False paths |\n\n### Feedback\n\n| Tool | Description |\n| ------------------------ | ------------------------------------------ |\n| `list_feedback` | List feedback for a project, newest first |\n| `get_feedback` | Get a feedback submission with its replies |\n| `update_feedback_status` | Update feedback status |\n\n### GitHub\n\n| Tool | Description |\n| -------------------- | ---------------------------------------------------------------------- |\n| `list_pull_requests` | List pull requests for a GitHub repository |\n| `get_pull_request` | Get details for a GitHub pull request (title, author, branches, stats) |\n| `comment_on_pr` | Post a comment on a GitHub pull request |\n\n### Interviews\n\nStructured user interviews: actors or generated personas run through a discussion guide (sections + questions/tasks) against a target URL, then Stunt Double synthesises themes and recommendations.\n\n| Tool | Description |\n| ----------------------------- | ------------------------------------------------------------------------------------------ |\n| `list_interviews` | List interviews in a workspace, optionally filtered by project |\n| `get_interview` | Get an interview with its discussion guide (sections + items) and participants |\n| `create_interview` | Create a new interview in a project (name, target URL, research brief) |\n| `update_interview` | Update an interview's name, target URL, research brief, or status |\n| `add_interview_section` | Add a section to the discussion guide |\n| `add_interview_item` | Add a question or task to a section |\n| `add_interview_participant` | Attach a participant, either an existing actor or an ad-hoc `persona_spec` |\n| `get_interview_participant` | Get a participant including their full transcript from the run |\n| `get_interview_report` | Get the current synthesised report (summary, themes, recommendations, per-question rollup) |\n| `launch_interview` | Launch the interview round (async). Returns the trigger run ID |\n| `regenerate_interview_report` | Re-run synthesis on existing transcripts (async). Returns the trigger run ID |\n\n### Prompts\n\nMost MCP clients (Claude, Claude Code, Cursor) surface these as slash commands. Each is a self-contained recipe: which tools to call, in what order, and how to report back.\n\n| Prompt | Description |\n| --------------------- | ------------------------------------------------------------------------------------------------------------------------- |\n| `validate_design` | Validate a live design, prototype, or preview URL (Figma Make, Claude artifact, v0, staging) with AI personas |\n| `verify_change` | Verify a shipped or previewed code change by running an actor through the affected flows, optionally commenting on the PR |\n| `run_user_research` | Run a structured multi-persona interview study and synthesise themes and recommendations |\n| `triage_feedback` | Triage user feedback on a project: cluster it, reproduce issues with an actor, and update statuses |\n| `setup_guardrails` | Stand up checklists for critical flows plus a workflow that re-runs them on a schedule or on deploy/PR events |\n| `check_brand` | Audit a product against brand and tone-of-voice guidelines, flagging deviations with evidence |\n| `check_design_system` | Audit a live product against its design system (typography, colour, spacing, components) on rendered pages |\n| `check_compliance` | Check a product against legal and compliance requirements and collect evidence for counsel to review |\n| `check_continuity` | Check continuity across surfaces (pricing, terminology, promises) between marketing, product, docs, and emails |\n| `stuntdouble_guide` | Orientation for Stunt Double: what it does, when to reach for it, and the full tool catalogue |\n\n### Resources\n\nRead-only context a client can attach without calling a tool. The guide and the connection are always listed; the workspace resources need `mcp:read`, the same as the tools that return that data.\n\n| URI | MIME type | Description |\n| -------------------------------------------------- | ------------------ | --------------------------------------------------------------------------------------------- |\n| `stuntdouble://guide` | `text/markdown` | What Stunt Double does, when to reach for it, the full tool catalogue and how to poll a run |\n| `stuntdouble://connection` | `application/json` | Who the connection acts as, the scopes it holds, the tools each unlocks and any it lacks |\n| `stuntdouble://workspaces` | `application/json` | The workspaces the connection can reach, with your role (same data as `list_workspaces`) |\n| `stuntdouble://workspaces/{workspace_id}/projects` | `application/json` | The live projects in one workspace, most recently opened first (same data as `list_projects`) |\n\n`resources/list` includes one projects entry per workspace, so a client can browse them without expanding the template.\n\n### Stunt Double Index\n\nThe [Stunt Double Index](https://index.stuntdouble.io) is a public ranking of how AI agents experience websites: each tracked domain is scored out of 100 from HTTP probes plus live agent sessions, one per AI provider and benchmark task. Index data is public rather than workspace data, so the read tools reach any tracked site.\n\n| Tool | Description |\n| ---------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ |\n| `get_index_report` | A site's score, band, rank, category and provider scores, frictions, failing probe checks and when it was last scored |\n| `list_index_sessions` | The agent sessions behind a score (newest run by default), with status, score, evidence, a summary and frictions. Filter by provider or category |\n| `search_index_domains` | Find a site by domain or name, or browse the leaderboard, optionally by sector |\n| `request_index_rerun` | Re-score a domain you own (`mcp:run`). Fresh probes now, about 24 agent sessions over a few minutes |\n\n`get_index_report`, `list_index_sessions` and `request_index_rerun` take a `domain`, or a `project_id` to use the Index domain linked to that project (which checks your membership like any project read). Only the domain's owner can re-run it. After a re-run, poll `list_index_sessions` with the returned `run_id` until the sessions finish, then read `get_index_report`.\n\n## Example prompts\n\nFour prompts that exercise the core of the server once it is connected:\n\n1. **Verify a flow:** \"Create a checklist that completes the checkout on https://checkout.demo.stuntdouble.io and reaches the confirmation page, then run it and tell me which checks failed.\"\n2. **Run a user interview:** \"Set up an interview with three personas (a first-time diner, a regular and a screen reader user) about booking a table on https://parlour.demo.stuntdouble.dev, launch it, and summarise the report.\"\n3. **Check agent readiness:** \"How well can AI agents use stripe.com according to the Stunt Double Index, and which categories are dragging its score down?\"\n4. **Triage feedback:** \"Summarise the open feedback on my main project, group it into themes, and mark anything already fixed as resolved.\"\n\n### Demo targets\n\nTry the server without pointing it at your own product. These public demo apps have deliberate defects for actors to find, so runs never touch real customers:\n\n| App | What |\n| -------------------------------------------------------------------- | ----------------------------------------------------------------------- |\n| [checkout.demo.stuntdouble.io](https://checkout.demo.stuntdouble.io) | Trailhead Outfitters: a one-page checkout |\n| [flights.demo.stuntdouble.io](https://flights.demo.stuntdouble.io) | Skyline Fares: one-way flight search, results, booking and confirmation |\n| [parlour.demo.stuntdouble.dev](https://parlour.demo.stuntdouble.dev) | Ondine, on Parlour: a restaurant booking flow with a deposit |\n\n## Privacy Policy\n\nThe server is hosted by Stunt Double and acts as the signed-in user. It reads and writes only the workspaces that user belongs to, within the OAuth scopes they grant (`mcp:read`, `mcp:write`, `mcp:run`). It does not read your conversation with the AI client beyond the arguments passed to each tool call, and it does not access the client's memory, chat history or files.\n\nData created through the server (projects, actors, checklists, runs, interviews, feedback) is stored in your Stunt Double workspace and handled under the [Stunt Double Privacy Policy](https://www.stuntdouble.io/privacy), which covers collection, use, storage, sub-processors, retention and your rights. Revoke a connection at any time by disconnecting it in your AI client.\n\n## Support\n\n- Email: [support@stuntdouble.io](mailto:support@stuntdouble.io)\n- Help centre and docs: [stuntdouble.io/support](https://www.stuntdouble.io/support)\n- MCP reference: [stuntdouble.io/support/docs/api/mcp](https://www.stuntdouble.io/support/docs/api/mcp)\n- Security reports: see [SECURITY.md](./SECURITY.md)\n\n## Transport\n\nThis server uses [Streamable HTTP](https://modelcontextprotocol.io/specification/2025-03-26/basic/transports#streamable-http) transport. The endpoint is:\n\n```\nhttps://app.stuntdouble.io/api/mcp\n```\n\n## MCP Registry\n\nThe server is listed in the official [MCP Registry](https://registry.modelcontextprotocol.io) as `io.stuntdouble/mcp-server`, so registry-backed clients and directories can find it by name:\n\n```bash\ncurl \"https://registry.modelcontextprotocol.io/v0.1/servers?search=io.stuntdouble/mcp-server\"\n```\n\nEach version merged to `main` is published there automatically (see [CONTRIBUTING.md](./CONTRIBUTING.md#releasing)).\n\n## Links\n\n- [Website](https://stuntdouble.io)\n- [Documentation](https://www.stuntdouble.io/support/docs)\n- [MCP server reference](https://www.stuntdouble.io/support/docs/api/mcp)\n- [Privacy Policy](https://www.stuntdouble.io/privacy)\n- [Terms of Service](https://www.stuntdouble.io/terms)\n- [llms.txt](https://www.stuntdouble.io/llms.txt)\n\n## Verifying changes\n\nFrom the repo root:\n\n```bash\nnode scripts/validate-json.mjs\nnpx --yes prettier@3.4.2 --check README.md CONTRIBUTING.md SECURITY.md CHANGELOG.md mcp.json .mcp.json server.json .cursor-plugin/plugin.json .cursor-plugin/marketplace.json .claude-plugin/plugin.json .claude-plugin/marketplace.json\n```\n\nMore context in [CONTRIBUTING.md](./CONTRIBUTING.md). GitHub Actions runs the same checks on push and pull requests.\n\n## License\n\nMIT\n"}