integration · linear

Claude + Linear — issue tracking, agent-shaped.

Q: Which OAuth scopes does wmcp.sh request?

read + write — covers all standard CRUD on issues, projects, comments, and labels. Linear doesn't support more granular scoping at the OAuth App level today.

Linear's API is GraphQL-only. wmcp.sh wraps the common operations as MCP tools so your agent can file bugs, comment on issues, move things across columns, and query projects without writing GraphQL. Connect Linear once; the worker injects your OAuth token on every call.

Last updated 2026-05-28 · works with Claude, OpenAI, LangChain, and any MCP client

Positioning

vs. the alternatives

Capability	@linear/sdk	Composio Linear	wmcp.sh
Setup time	Install SDK + write GraphQL	Platform + per-customer OAuth	Connect Linear once
GraphQL exposure	Full — you write queries	Abstracted via REST-ish methods	Curated MCP tools (no GraphQL knowledge needed)
Auth	PAT or OAuth, you store	Platform manages	OAuth vault in wmcp.sh
MCP shape	Wrap each query yourself	Auto-mapped	Native MCP tools
Webhooks	Yes	Platform → workflow	Out of scope for v0
Cost	Free SDK	Platform tier	Free 100/day + $29/mo Pro

What you get

Tools your agent calls

All MCP-shaped — consumed directly by Claude tool_use, OpenAI function-calling, or LangChain.

Tool	Type	Returns
`issueCreate`	Live action	New issue with ID + URL
`issueUpdate`	Live action	Updated issue
`issuesByTeam`	Live action	Filtered issue list
`issueSearch`	Live action	Issues matching query
`commentCreate`	Live action	New comment ID
`projectsByTeam`	Live action	Project list
`workflowStates`	Live action	State machine for a team
`myIssues`	Live action	Issues assigned to current user

Integrate

Python — agent files a bug from a deploy log

from wmcp import WmcpClient
from wmcp.anthropic import to_anthropic_tools, execute_tool_use
from anthropic import Anthropic

client = WmcpClient(api_key="webmcp_live_…")
spec   = "https://wmcp.sh/api/v1/curated/linear"   # curated tool list, not raw GraphQL
tools  = client.tools(spec)

anthropic = Anthropic()
msg = anthropic.messages.create(
    model="claude-opus-4-7",
    max_tokens=1024,
    tools=to_anthropic_tools(tools),
    messages=[{"role": "user",
               "content": "File a P1 bug in the BACKEND team: '/u/<hash> 500s on cold cache when seen: metadata missing'."}],
)

for block in msg.content:
    if block.type == "tool_use":
        # Worker injects your connected Linear OAuth token
        result = execute_tool_use(client, spec, block.model_dump())
        print(result)

FAQ

Frequently asked

Does this support arbitrary GraphQL or curated tools only?

Both. The curated tool list (above) covers ~80% of agent use cases without exposing GraphQL. For anything else, hit Linear's GraphQL endpoint directly via the OpenAPI-style passthrough — wmcp.sh injects auth, you craft the query.

Which OAuth scopes does wmcp.sh request?

read + write — covers all standard CRUD on issues, projects, comments, and labels. Linear doesn't support more granular scoping at the OAuth App level today.

Team / project scoping?

OAuth tokens are workspace-scoped. Within a workspace, the agent can act across any team the OAuth user has access to. Pass teamId in tool args to scope writes to a specific team.

Rate limits?

Linear's API has both per-user and per-IP rate limits (1500 / hour typical). wmcp.sh respects 429 responses. For bulk operations, batch via Linear's native batching where supported.

Webhook support?

Not in v0. wmcp.sh covers the request-side surface. If you need to receive Linear events, set up a separate webhook endpoint and have it call wmcp.sh tools in response.

API agent-readiness

Linear's GraphQL + OAuth is a model for shipping an API agents can drive. See the API-side checklist.

→ /agent-ready/api · cornerstone · done-for-you ($499+)

Need this done for you?

Skip the wiring — we build, deploy, and monitor.

Custom adapter + hosted MCP at mcp.yourbrand.com + verified badge. From $499 one-time setup.

See /managed → Submit (free)