generated: '2026-07-19' method: derived source: openapi/langai-api-openapi.yml status: candidate note: Lang.ai publishes no official hosted or remote MCP server — no server is listed in the docs, the lang-ai GitHub organization, or the public MCP registries, and the company is now part of Capacity. The tools below are a candidate surface derived one-to-one from the documented REST operations, offered as a starting point for anyone wrapping the Lang.ai API for agents. Do not present this as a Lang.ai-published server. server: name: langai transport: stdio url: null auth: type: bearer env: LANGAI_API_TOKEN note: Also requires the tenant subdomain, e.g. LANGAI_INSTANCE=company for https://company.lang.ai/api/v1. tools: - name: create_project description: Create a Lang.ai classification project from a CSV dataset of customer interactions, naming the text column and the teams that may access it. source_operation: openapi/langai-api-openapi.yml#createProject consequence: write - name: list_projects description: List the classification projects on this Lang.ai instance with their id, name, status and creation date. source_operation: openapi/langai-api-openapi.yml#listProjects consequence: read - name: get_project_tags description: Get a project and the list of tags defined on it, including whether each tag is a draft. source_operation: openapi/langai-api-openapi.yml#getProjectTags consequence: read - name: analyze_document description: Classify a piece of text against a project and return the matched tags and the intents/features detected, without persisting the document. source_operation: openapi/langai-api-openapi.yml#analyzeDocument consequence: read - name: save_document description: Save a document into a project with optional metadata and date, returning its classification. Supplying an existing document id overwrites that document. source_operation: openapi/langai-api-openapi.yml#saveDocument consequence: write deployment: mode: none verified: derived tools: 5 checked: '2026-08-12' source: catalog MCP census