Clapr MCP server
Clapr exposes a remote MCP (Model Context Protocol) server so AI agents can read and manage your content — boards, cards, scripts, comments, performance. Free on every plan.
- Endpoint:
https://www.clapr.io/api/mcp(Streamable HTTP) - Auth: personal access token via
Authorization: Bearer …, or OAuth 2.1 for clients like claude.ai — no key to paste. - A token grants access to all your workspaces. Writes respect your role: owners and editors can change things, viewers are read-only.
1. Generate a token
In Clapr, open Settings → API & AI tools, enter a name and click Generate token. Copy it immediately — it is shown only once. You can revoke it anytime from the same screen.
2. Connect your client
Claude Code
claude mcp add --transport http clapr https://www.clapr.io/api/mcp \
--header "Authorization: Bearer clapr_mcp_xxx"claude.ai (web) — no token needed
Add a custom connector with the URL https://www.clapr.io/api/mcp. claude.ai auto-discovers Clapr's OAuth server, opens a Clapr sign-in and consent screen, and gets a token automatically (Authorization Code + PKCE, Dynamic Client Registration).
Codex, Cursor & other MCP clients
Add a remote / HTTP MCP server with:
- URL:
https://www.clapr.io/api/mcp - Header:
Authorization: Bearer clapr_mcp_xxx
Use https://www.clapr.io/… — the apex clapr.io redirects to www.
3. Tool reference
| Tool | Access | What it does |
|---|---|---|
list_workspaces | read | All workspaces you can access, with your role in each. |
list_statuses | read | Board columns (statuses) of a workspace, in order. |
list_cards | read | Content cards, filterable by workspace and status. |
get_card | read | One card in full: script, references, comments, performance. |
get_performance | read | Performance snapshots (views, likes, comments) over time. |
create_card | write | Create a card: title, description, script, scheduled date, target column. |
update_card | write | Edit a card's fields. |
move_card | write | Move a card to another column. |
add_reference | write | Attach a reference URL to a card. |
add_comment | member | Leave a comment on a card (any member, including viewers). |
read — any member. write — owner or editor. member — any member including viewers.
Try it
Once connected, ask your agent things like:
- “Draft my next 3 video ideas as cards and schedule them for next week.”
- “Move the n8n video to Editing and leave a comment for my editor.”
- “Which of my last 10 videos performed best? Suggest two follow-ups.”
Agent-created cards land in your first board column with an Agent badge, so you and your editor always see what came from an agent and review it before it moves on. See docs for agents for recommended conventions.
Security
- Only the SHA-256 hash of each token is stored; the raw token is shown once.
- Tokens are revocable and track
last_used_at. - All access is scoped to your workspaces and enforced server-side, per role.
Prefer plain HTTP? Use the REST API — same tokens, same permissions.