vje013

vje013 / darwin-agentic-cloud

Public

Verifiable and free compute for AI agents. Cryptographically signed execution attestations. MCP-native.

darwin-agentic-cloud.fly.dev agent-infrastructure agent-skills agent-wallet agentic-cloud ai-agents
11
0
89% credibility
Found May 29, 2026 at 13 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Darwin Agentic Cloud is an open-source platform that provides free, verifiable compute for AI agents. When an AI agent needs to execute code, Darwin runs it and returns a cryptographically signed receipt (called an attestation) that proves exactly what happened: what code was executed, what output was produced, which computer handled it, and how much it cost. Anyone can verify these receipts against a public keylist without trusting Darwin itself. The platform supports multiple execution backends (local computer, cloud services) and automatically picks the cheapest option. It also integrates as a plugin for popular AI assistants, giving them verifiable code execution capabilities.

How It Works

1
💡 You hear about a way to run AI agent code for free

A friend tells you about Darwin - a service that runs your AI agent's code and gives you a signed receipt proving it happened.

2
📦 You install Darwin in seconds

With one simple command, Darwin is ready on your computer. It automatically finds what compute options you have available.

3
🚀 You ask your AI agent to run some code

Your AI agent wants to execute code - maybe analyze a CSV, fetch data, or run a calculation. It calls Darwin to do the work.

4
📜 You get back a signed receipt

Darwin runs the code and returns a beautiful certificate proving exactly what happened: what code ran, what output came out, which computer handled it, and how much it cost.

5
✅ Anyone can verify the receipt is real

The receipt has cryptographic signatures. A regulator, auditor, or another AI can check the public keylist and confirm the execution actually happened.

6
You connect Darwin to your AI assistant
🤖
Claude Desktop

Install the Darwin plugin and your AI assistant gains verifiable code execution.

💻
Cursor

Add Darwin to Cursor and every line of code your AI writes comes with proof.

🎉 Your AI agent now has verifiable, auditable execution

Every action your AI takes leaves a paper trail. You can prove to anyone that the right code ran, producing the right output, at the right cost.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 13 to 11 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is darwin-agentic-cloud?

Darwin Agentic Cloud is a Python tool that gives AI agents a cryptographic receipt for every task they execute. When your agent runs code, Darwin signs the result with an Ed25519 key, proving what ran, where, for how long, and at what cost. The key innovation is the dual-signature model: the execution substrate signs its identity, then your local machine signs off on the result. Anyone can verify an attestation against the public keylist without trusting Darwin itself.

The project ships as a pip-installable CLI and Python library. You get seven commands: `run` to execute code with a signed receipt, `verify` to check signatures cryptographically, `price` to preview costs before running, `list` to see available backends, `try` for offline-only execution, `who` to inspect signers, and `sign` for key ceremony. It auto-discovers execution environments (local Docker, AWS Lambda, AWS Batch, Modal) and routes to the cheapest unless you override.

Why is it gaining traction?

The use case is concrete: if your agent books travel or buys office supplies, you need proof it actually happened. Darwin solves the verification gap that every agent framework ignores. The "value-added service" block in every receipt tells you the cost cap, actual cost, headroom, and routing decision, so audit trails are self-contained.

It is also genuinely free, no credit card required, with a revenue model built on merchant referrals rather than compute charges. The MCP server integration means agents built on Claude Desktop or Cursor can call Darwin as a native tool. The signing chain is auditable end-to-end, and the schema is public.

Who should use this?

Agent developers building automated purchasing flows, procurement automation, or any workflow where spending authorization and auditability matter. Developers in regulated industries (finance, healthcare, procurement) who need verifiable execution records for compliance. Anyone running agentic workflows that must satisfy auditors, counterparties, or internal policy.

Not yet suitable for production without careful review: the project is alpha status with a credibility score of 0.9% and only 11 stars.

Verdict

A compelling idea at a very early stage. The code is structured professionally with typed Python, comprehensive CLI, and multiple execution backends. The verification path is genuinely open source. If you are building anything where "prove this agent actually ran this code" matters, watch this. Do not bet production workloads on it yet, but the problem it solves is real and underserved.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.