atlas-mcp
Connect AI coding agents to Atlas knowledge hub for context retrieval and work capture.
README
atlas-mcp
Connect any AI coding agent (Cursor, Claude Code, Windsurf, VS Code/Copilot, Cline…) to your team's Atlas knowledge hub. The agent pulls your approved prompts, docs and scoped secrets as context — and Atlas passively captures what the AI builds, so the team can see it without anyone using Jira.
Install globally (once per machine)
bun install -g github:TheDivyanshShukla/atlas-mcp
# or: npm install -g github:TheDivyanshShukla/atlas-mcp
This puts atlas on your PATH — repo setup, hooks, and the stdio MCP server IDEs spawn.
One-command setup
# in your repo root
atlas .
# or: atlas . --project <atlas-project-slug>
This writes a .atlas marker (safe to commit) and adds the Atlas MCP server to every agent config it
finds (.mcp.json for Claude Code, ~/.cursor/mcp.json, Windsurf, .vscode/mcp.json).
Then create an MCP key in Atlas → Settings → API keys → MCP key (pick toolsets, read-only, expiry), and export it:
export ATLAS_MCP_KEY=atlas_mcp_xxxxxxxx # add to your shell profile
export ATLAS_BASE_URL=https://atlas.naravirtual.in
Claude Code only — capture every prompt + session automatically (zero agent cooperation):
atlas hooks install
What it exposes (scope-gated — tools you can't use are hidden)
- context:
atlas_context(one-call prompts+docs bundle),atlas_search,atlas_list_prompts,atlas_get_secret(audited decrypted .env values — use instead of pasting creds). - capture:
atlas_log_work(records summary + prompt + changed files),atlas_save_prompt. - tasks:
atlas_list_tasks,atlas_create_task.
How it works
A thin, cached client over the Atlas REST API. Reads are served from a local cache
(~/.atlas/cache, stale-while-revalidate) so they're instant and work offline; writes queue and flush
when the network returns. All auth, fine-grained scopes and audit are enforced server-side.
The .atlas file binds the repo to an Atlas project and selects toolsets:
{ "project": "general", "autoCapture": true, "toolsets": ["context","capture","tasks"], "readOnly": false }
The key (secret) is never in .atlas — it lives in your IDE env / shell profile.
Agent transport options
Repo: https://github.com/TheDivyanshShukla/atlas-mcp
Remote (zero local install beyond the key) — point any client at the hosted server:
claude mcp add --transport http atlas https://atlas.naravirtual.in/api/mcp --header "Authorization: Bearer $ATLAS_MCP_KEY"
Or for Cursor/Windsurf/VS Code, paste the remote config (see Atlas → Connect agent for your exact snippet).
Local (repo-aware, recommended for capture) — stdio via globally installed atlas (reads .atlas, offline cache):
atlas . # set up a repo
# agents are configured to run: atlas
Auth is read from ATLAS_MCP_KEY (set once in your shell profile). Default endpoint is
https://atlas.naravirtual.in; override with ATLAS_BASE_URL.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.