Skip to content

langchain-linear-writeback

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

langchain-linear-writeback

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
src
src
 
 
README.md
README.md
 
 
package-lock.json
package-lock.json
 
 
package.json
package.json
 
 
tsconfig.json
tsconfig.json
 
 

README.md

LangChain × Linear writeback (via Relayfile)

LangChain (via @langchain/langgraph createReactAgent) creates / updates / deletes Linear labels through Relayfile. Schema-validated, with comprehensive op-status proof (writeback.provider=linear → op succeeded → real Linear UUID in providerResult.externalId).

Quickstart

agent-relay cloud login
cd examples/integrations/langchain-linear-writeback
npm install
CLOUD_WORKSPACE_ID=<your-app-uuid> npm run smoke

Then for the interactive agent:

ANTHROPIC_API_KEY=sk-ant-… npm run dev

For CI: set CLOUD_API_ACCESS_TOKEN (and optionally CLOUD_API_REFRESH_TOKEN, CLOUD_API_URL).

Important

Every smoke run creates ONE real Linear label and deletes it before exit. Test labels are prefixed relayfile-writeback-test. See vercel-ai-sdk-linear-writeback/README.md for the full contract details.

Why this vs. a provider MCP?

Honest answer in packages/agents/README.md: the MCP wins for one-shot single-provider work; @relayfile/agents is structurally better as integrations and agents compound.