Darwin Agentic Cloud MCP server
Commit Oct 20, 2026Provides verifiable, sandboxed cloud compute for AI agents with signed receipts for every autonomous action and cost.
Also known as vje013/darwin-agentic-cloud
- Stars
- 37
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- —
- Last commit
- Oct 20, 2026
- License
- Apache-2.0
What is Darwin Agentic Cloud MCP?
Darwin Agentic Cloud provides verifiable, sandboxed cloud compute for AI agents with signed receipts for every autonomous action and cost.
Summary generated from the project's README
About Darwin Agentic Cloud MCP server
Providing sandboxed cloud compute with signed receipts for autonomous AI agents is the primary function of this server. It allows developers to run code in a secure environment where every action, choice, and purchase is bound and signed. This ensures that agents only operate within authorized limits and provide a verifiable audit trail for their activities.
Users can deploy this locally using Docker or connect it to MCP clients like Claude Desktop or Cursor via the darwin mcp install command. The system requires a running Docker daemon for local execution. It uses webMCP and Docker to manage tasks, allowing you to set specific constraints such as a cost cap or a timeout for any given job.
Every execution produces a signed receipt that binds the exact code, output, substrate, and cost. You can use the darwin verify command to check these signatures offline against a public keylist to ensure no data has been altered. This setup is designed for agent builders who need to prove what an agent did and how much it spent.
What you can do with it
- 1Execute sandboxed Python codeRun a specific Python script or command within a secure environment that tracks all outputs and costs.
- 2Audit autonomous agent actionsVerify the integrity of an agent's history by checking signed receipts against public keys.
- 3Control agent spending limitsRestrict an AI agent's financial impact by enforcing a maximum cost for its computational tasks.
- 4Integrate verifiable compute into IDEsAdd Darwin Agentic Cloud tools to Cursor or Claude Desktop to give your coding agents a receipt for every step.
Written from the project's README. Every use case is checked against a quote from it.
Automated checks
Ran Nov 14, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachablevje013/darwin-agentic-cloud is public and reachablegithub.com/vje013/darwin-agentic-cloud ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)Apache-2.0spdx.org/licenses/Apache-2.0.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-10-20github.com/vje013/darwin-agentic-cloud ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubDarwin Agentic Cloud
Verifiable and free cloud compute for AI agents.
We provide an agent-first environment where every autonomous action and choice is bound and signed, including what they buy.
Your agent acts only where it can provably succeed, spends only what you authorized, and brings back a signed receipt for every step it took.
No new accounts. No new API keys.
Open source. Free to users. All agent and human contributors welcome.
Built on: webMCP + Docker
Install: pip install darwin-agentic-cloud
Live demo: https://darwin-agentic-cloud.fly.dev/demo
Public keylist: https://darwin-agentic-cloud.fly.dev/.well-known/substrate-keys.json
v0.2 Schema: https://darwin-agentic-cloud.fly.dev/.well-known/schemas/attestation/v0.2
Darwin's Mission
Increase AI adoption by increasing AI safety.
How to Run
1. Install and start Docker.
pip install darwin-agentic-cloud
darwin --help # confirms the install and lists every command
open -a Docker # macOS. On Linux: sudo systemctl start docker
docker info # should print server details, not an error
Local runs need a running Docker daemon.
2. Run something and save the receipt.
darwin run 'print("hello")' --cost-cap 0.10 --timeout 60 --save receipt.json
The job runs sandboxed under a hard cost cap and timeout. You get the output plus a signed receipt binding the exact code, output, substrate and cost. To run a script, pass its path: darwin run ./my_task.py ...
3. See who signed it.
darwin who receipt.json
Shows the substrate and operator keys behind the receipt. Keep receipts and you have an audit trail of every agent action.
4. Verify it.
darwin verify receipt.json
Anyone can check the signature offline against the public keylist. Change one byte and it fails.
No Docker? Verify a hosted receipt instead.
curl -s https://darwin-agentic-cloud.fly.dev/demo/attestation.json -o att.json
darwin who att.json
darwin verify att.json
For agent builders: darwin mcp install adds DAC as tools in Claude Desktop, Cursor or any MCP client, so your agent gets a receipt for every step.
Darwin Agentic Cloud MCP questions
How can I use Darwin Agentic Cloud with my AI agent?
You can use the command darwin mcp install to add DAC as tools in Claude Desktop, Cursor, or any MCP client.
Does running tasks locally require Docker?
Yes, local runs require a running Docker daemon.
Can I verify the authenticity of a task receipt?
Yes, anyone can verify the signature offline using the public keylist, and any modification to the file will cause verification to fail.
How do I limit the cost and time of a task?
You can use the --cost-cap and --timeout flags when running a command.