Glean Developer Platform

Your Coding Agent Knows Code. Glean Teaches It Your Company.

Code assistants can write fast. Glean gives them the context they’re missing — docs, decisions, people, tickets, and code — so they can build in the real world, not in a vacuum.

• Ground plans in reality – Pull in design docs, prior decisions, and team context before writing code
• Works where you work – Use Glean with Claude Code, Cursor, and other MCP-compatible tools
• Permission-aware from day one – Every query respects your company's access controls

Connect Glean to Your IDE

Connect Any Agent to Your Company's Knowledge

Not every agent will run on Glean. That’s fine. Glean is the knowledge layer that connects any agent to your enterprise context through one secure MCP endpoint.

• One connection, many hosts – Works with Claude, Cursor, Copilot, ChatGPT, Windsurf, and 20+ MCP hosts
• Framework-agnostic – Plug Glean into LangChain, OpenAI Agents SDK, Google ADK, or any MCP-compatible framework
• Bring agents to the data – Expose Glean agents as tools anywhere your developers work

Get Started with MCP

Build Agents That Actually Know Your Business

Glean agents reason over your company’s knowledge graph, not just a prompt window. Search, analyze, find experts, and take action with enterprise-safe access built in.

• Multi-step by design – Plan, execute, and iterate across complex workflows
• Grounded in company context – Search, chat, code, people, and meetings as native tools
• Deploy where you need them – Run on Glean, call via API, or expose as MCP tools in any host

Choose Your Agent Approach

import os
from crewai import Agent, Crew, Task
from glean.agent_toolkit.tools import glean_search

os.environ["GLEAN_API_TOKEN"] = "your-api-token"
os.environ["GLEAN_INSTANCE"] = "your-instance-name"

researcher = Agent(
    role="Research Specialist",
    goal="Find company docs and information",
    tools=[glean_search.as_crewai_tool()],
)

crew = Crew(agents=[researcher], tasks=[
    Task(description="Find remote work policy", agent=researcher),
])
result = crew.kickoff()

Bring Every Data Source into the Graph

The better your knowledge graph, the better every search, agent, and integration performs. Glean connects 100+ apps out of the box and lets you index everything else through flexible APIs.

• Bulk or real-time – Upload millions of documents or sync continuously
• Flexible schema – Model metadata, relationships, and custom properties your way
• Foundation for everything above – A richer graph makes every agent smarter

View Indexing APIs

import os
from glean import Glean, models

with Glean(
    instance='acme',
    api_token=os.getenv('GLEAN_INDEXING_TOKEN', ''),
) as g:
    document = models.DocumentDefinition(
        id='doc-123',
        title='Q4 Sales Report',
        body='Our Q4 performance exceeded expectations...',
        datasource='internal-docs'
    )

    g.indexing.index_document(document=document)

Connect your AI coding assistant to Glean's developer documentation via MCP for instant access to API references and guides. Get started.

---

Integrate with AI Tools

Connect Claude Code, Cursor, Copilot, and any MCP host to your enterprise knowledge.

Build with Our APIs

Search, chat, and agent APIs with client libraries for Python, TypeScript, Java, and Go.

Connect Your Data

Bring any source into Glean with our connector framework and indexing API.