Every project, set up for agents
Every project gets BMAD, the ADE methodology and eight core agents, plus Solidity, Solana, Go or Rust agents for its stack.
How project setup works
Verified, not vibed.
The terminal built around Claude Code, with Codex, Gemini and local models alongside it, and a methodology that decides when the work is actually done.
Features
AI agents and blockchain tooling, in one window.
Run AI agents well. Ship smart contracts safely.
01
The IDE
Any CLI agent, side by side, in projects set up for them, with every token accounted for.
Claude Code, first class
Every project gets sub-agents in .claude/agents/ and a CLAUDE.md constitution. The fleet gives each sub-agent its own lane, spend is read from Claude Code's own transcripts, and ADE ships a plugin for it with 7 agents and 13 skills. The others run beside it; this is the one it is built around.
Every agent on one fleet view
Every agent and sub-agent on one live timeline, with its real cost.
66 agents, one keystroke away
47 role-and-domain profiles and 19 standalone roles. Backend, Web3, DevOps, testing and more.
A scratchpad for prompts
Draft, chain and reuse prompts. Drafts survive crashes.
Know the cost, spend less
Cost, cache savings and context per session, with one-click savers.
02
Built for serious software engineering work
An agent will tell you it is finished. The question is who checks. ADE sets every project up with a methodology that answers that, and roles that cannot mark their own homework.
Done is a command, not an opinion
Every project agrees one verification command up front. A story is Done when it passes on that commit, with the output shown. Not when an agent says so.
No agent certifies its own work
The Developer writes the failing test first, then the code. QA judges the result and may not edit code or tests, because a verifier that repairs what it measures can make its own verdict come true.
An adversarial reviewer that cannot patch
It ships with a read-only tool allowlist on purpose: Read, Grep, Glob, Bash. It judges and reports. Widening it to unblock a story is how review stops meaning anything.
Nineteen roles, each with boundaries
Analyst through security engineer, SRE and release manager. Each one states what it owns, what it must not touch, and whose job the rest is, so parallel agents stop contradicting each other.
BMAD in every project, and actually used
75 files of tasks, checklists, templates and workflows. Eight roles name the ones that do their job, so the Scrum Master shards with create-next-story and QA traces criteria with trace-requirements instead of improvising both.
What they learn stays in your repo
A knowledge store per role, committed with the project. Shared contracts go to the context store where parallel agents read them. An update never overwrites either.
03
Agentic workflow and blockchain engineering
The two places unreviewed code costs the most. Architects for agent design, RAG and LLMOps sit beside Foundry, Hardhat and Anchor. Smart contracts run from first test to reviewed deploy, and ADE never touches private keys.
One key for the whole toolchain
Build, test, gas, coverage, Slither and a local chain.
A workbench for tests
Each test's gas, fuzz runs and counterexamples. Deploy and call on Anvil.
Reviewed deploys, readable ABIs
Deploys are typed for you to review, never run. ABIs with selectors.
Architects for agents and for chains
AI agent, RAG, LLMOps and workflow automation architects, beside DeFi, tokenomics and cross-chain. Plus an AI × Web3 architect for where they meet.
New
Project setup, Anti-slop, MCP library and Secrets vault.
Free and open source
Every feature. No licence, no account, no telemetry.
ADE is MIT licensed. Use it at work, fork it, ship your own build, no permission needed.
The whole app
There is no paid tier, because there is only one tier.
Free forever
- Fleet view, tokens panel, scratchpad, editor and the contract workbench
- macOS (Apple Silicon and Intel) and Linux
- Your keys and secrets stay in your OS keychain, never synced to us
The source
Read it, change it, or lift the parts you want.
MIT licensed
- The whole app: Tauri and Rust backend, React frontend
- Build it yourself and run your own version
- Issues and pull requests welcome
The agents
The nineteen role definitions live in their own repo.
Yours to fork
- Edit a role once and use it across every project
- Versioned and released, so a project can pin what it uses
- Share a house style with your team
Used on real work
Not a demo. The tool its author ships with.
ardata.tech
Visit ardata.tech
The tool behind ardata's projects
ADE is what ardata builds client work in. The agent fleet, the methodology and the contract tooling all exist because the work needed them.
caiden.tech
Visit caiden.tech
The tool behind CAIDen
AI agents for voice, SMS and chat, built and deployed for each customer. Booked, NarraLearn and the Healthway front desk were all built in ADE.
ardata academy
Explore the training
The tool behind ardata academy's AI training
Learn the workflow in ardata academy's course, then use the same tool on real projects.
Shipped with it
Products built using ADE
Not a feature list. Things in production, built by the people who build ADE.
W3
The operating system for autonomous finance: build and control agent-powered financial workflows.
KONEKSI
Distributed-storage disaster recovery for files and databases, with automatic backup and verified restores.
DatArk Managed
Restores a real sample every month and publishes the verification report, rather than a backup-success log.
DeltaVault
Data disaster recovery that runs on-premises or in the cloud, complementing high availability.
CAIDen
AI agents for voice, SMS and chat, built and deployed per customer.
Healthway Front Desk
Answers patients on chat, Viber and the phone, and books their appointments.
Booked
One QR code across yard signs, reels, flyers and cards, turning every one of them into a booking.
NarraLearn
Turns a curriculum into a voice-narrated course with an AI tutor built in.
Crimson Security
Canadian cybersecurity firm: compliance, penetration testing and monitoring.
Wodan AI
Zero-trust AI.
Goldfish (GGBR)
A gold-backed token, over-collateralised beyond 100% by audited reserves.
Hoop Culture
Runs amateur basketball leagues, scoring games live by voice.
arena.stats
Leaderboards, rankings and match stats across every sport and registered athlete.
AR Code Scan
Scans code from GitHub, Replit or Lovable for security holes and production readiness.
Savvy
A real-estate assistant that answers on properties, neighbourhoods and market conditions.
IPFSR
A pluggable IPFS datastore that replaces Kubo flatfs and replicates to Sia renterd.
zipfs
IPFS implemented in Zig.
And hundreds of proofs of concept and pilot projects.
FAQ
Questions, answered
Is it really free?
Yes. Download ADE; nothing is locked.
How does licensing work?
It does not. ADE is MIT licensed: no keys, no accounts, no seats to count. Use it at work, fork it, ship your own build.
Which agents does it support?
Any CLI agent runs here, but ADE is built around Claude Code and goes deepest with it: sub-agents in .claude/agents/, a CLAUDE.md constitution, sub-agent lanes on the fleet timeline, and real spend read from its transcripts. Codex, Gemini and Ollama run side by side and show no cost, because they publish none.
How many agents can run at once?
As many as your machine and your provider's rate limits allow. Ten is what we test.
Does it store my keys?
API keys stay in your OS keychain, never synced. ADE never handles private keys.
Which platforms does it run on?
macOS (Apple Silicon and Intel) and Linux (.deb or AppImage). On macOS you can also run brew install --cask alvin-reyes/tap/ade. Windows support is coming soon.
Build your next product in ADE
Free and open source. Every feature, no account.