btahir

btahir / agent-order

Public

The Order of the Agents is a council of AI agents (Codex, Claude) that deliberate to produce final plans, PRDs, and decisions.

15
2
100% credibility
Found Apr 28, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A tool that coordinates multiple AI agents to deliberate on scenarios and produce structured decision documents like product requirements, architecture records, and migration plans through independent planning, anonymous critique, revision, and synthesis.

How It Works

1
📖 Discover the helper

You find a clever tool that lets AI assistants debate your ideas to create clear plans and decisions.

2
🛠️ Get it ready

Follow easy steps to set up the tool on your computer so it's all prepared to go.

3
✏️ Describe your challenge

Write a short note about the decision or plan you need, like building a feature or choosing tools.

4
🚀 Start the agents' council

Choose a plan type like a product guide or decision record, and launch the group discussion.

5
👥 Watch them team up

The assistants think on their own, challenge ideas without knowing who said what, revise, and blend the best parts.

🎉 Receive your ready report

Enjoy a polished document with the final plan, what happened in the debate, and a shareable page to show others.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 15 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 agent-order?

Agent-order runs a council of AI agents like Codex and Claude that deliberate on your scenario to output polished PRDs, ADRs, RFCs, migration plans, or build-vs-buy memos. Feed it a markdown file or text prompt via CLI—like `agent-order prd ./scenario.md`—and get a final report, decision log, shareable HTML index, and full audit trail of every agent's position, critique, and revision. Built in TypeScript, it defines agent order meaning as independent plans, anonymized critiques, revisions, and synthesis, ensuring decisions beat single-model confidence.

Why is it gaining traction?

Unlike solo AI chats, it forces agent ordering where models challenge anonymized peers, preserving dissent in a minority report and surfacing the strongest ideas. Depths like quick (two agents) or deep (four with aggregators) control cost and rigor, plus human input on blockers and replay for evals make it practical for real workflows. Templates enforce structure (e.g., ADR rubric for reversibility), and GitHub integration via order github actions or copilot fits dev rituals without fluff.

Who should use this?

Tech leads drafting PRDs or RFCS, staff engineers on build-vs-buy or migration plans, PMs needing incident reviews. Ideal for frontend/back-end teams debating agent order valorant-style prioritization or order management.agent systems, not quick prototypes.

Verdict

Try it for decisions worth $5-20 in API calls—early at 15 stars and 1.0% credibility, but solid docs, evals, and mock mode lower risk. Pairs well with claude/codex CLIs; skip if single-model suffices.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.