Skip to content

About

Claude Code and Codex plugins by ArtyomITA - chatroom: put your agents in one room and watch them work (live group chat, task board, 3D voxel office)

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

266 Commits

Folders and files

Repository files navigation

chatroom: agents that have to prove it

Put your agents in one room and watch them work. Or hand them one file and a number to beat.

By Adrian Ciprian Iancu (GitHub · Hugging Face · LinkedIn · site)

chatroom Claude Code Codex papers license last commit stars

Give several coding agents the same task and they will agree with each other, call it done and move on. I wanted a team that argues, shows its evidence and cannot close until the proof is there, so I wrote those rules in code instead of asking nicely in a prompt.

$ echo "CLOSED v2" | chat.py post --as ALFA
error: PLAN v2 cannot be closed yet: no vote yet from BETA, GAMMA. Nothing was posted.

$ chat.py task done T1 --proof "looks fine" --as BETA
error: a proof must name what shows it works: a command and its exit code, tests passed, file:line, #message, a hash, a path or a link
(task done ID --proof "pytest -q: 12 passed, exit 0"). What you did not do goes in --gaps. Nothing was closed.

$ chat.py lock acquire bench --as BETA
lock bench is held by ALFA (29 min left), you are in queue at place 1: end your turn, the room tags you when it is yours

Two rooms

A room A research room
You give it A topic or a job One file, an evaluator, a number to beat
The team Agents you pick, with a lead Fixed roles: workers, proposer, auditor, registrar, searchers
The rules Plans, votes with evidence, tasks with proof of done, leases, locks Phases with gates, a sealed evaluator, a worktree per worker, repeated measures, a final run on held-out data
Who checks chat.py, the same for every agent A kernel that measures every attempt the same way and keeps the books, all night if you like

Both run in Claude Code and in Codex, and the two mix: a Claude Code room and a Codex room can share one office.

What you get to watch

  • A live group chat with no turns: agents write when they have something to say, and you write to all of them from the same page.
  • A 3D voxel office next to it, where you see who is coding, who is reading and who is waiting for you. It has a lounge, games, an arena and an arcade cabinet you can play with a friend.
  • A relay that wakes the lead only when something needs it, with the rest in one digest at the next wake.
  • chat.py cost: what each agent really spent, read from its transcript.

Lend your agents to a friend

Share the common room with one address and a 6-digit PIN, through a Cloudflare tunnel, Pinggy or your local network, with no account. Your friend crosses the bridge into your office, chats, plays with you and can ask for help. You approve in the panel, choosing model, effort and minutes, and a helper joins their chat and their office, where they can tag it and send it files. It works both ways, and between Claude Code and Codex: a Codex user can join your common room and lend or borrow agents. Only the common room leaves your computer.

Summon agents from a list of hosts

/spawn puts every free model from every provider with a free tier in one list: OpenRouter, NVIDIA NIM, Groq, Google AI Studio, Mistral, Cohere, Cloudflare, Hugging Face and more. Pick from cards with each model's scores and its 10 strongest benchmarks, give it a name and a mission, and it joins the room as a role.

A role outlives its model. When a host runs out of quota, the role moves to another host of the same model and keeps its mission, tasks and notes. When no host is left its name blinks and you pick the replacement. Keys stay encrypted outside the room and never reach the chat, the logs or a URL.

Install

In Claude Code:

/plugin marketplace add Partido-De-What/claude-plugins
/plugin install chatroom@artyomita-plugins

Then start a room with /chatroom:start <topic>.

In Codex:

codex plugin marketplace add Partido-De-What/claude-plugins
codex plugin add chatroom@artyomita-plugins

Then, in a new thread, ask Codex to open a chatroom on your topic.

Every mechanism is listed point by point in the chatroom README.

The papers behind it

These papers shaped the plugin, and each row is something they changed in the code. All are arXiv preprints.

chatroom

Paper What it changed
Destefanis et al., When Agents Coordinate: Measuring Coordination in Multi-Agent AI Coding Work goes through the board and files, with one-line return reports and no round of introductions
Yang et al., When Parallelism Pays Off: Cohesion-Aware Task Partitioning for Multi-Agent Coding Tasks are split along their declared files
Weinberger et al., Token Reduction Is Not Cost Reduction Compression stays off by default and never runs as a proxy; chat.py cost measures instead
Wang et al., Do We Need Complex Topology Control? Distinct-Peer Random Routing Improves Cost-Efficiency in Sparse Multi-Agent Debate Two votes at most on a plan, then the lead decides
Zerhoudi et al., The Compaction Cliff in Long-Running AI Agent Memory notebook --rule: rules kept word for word through a compaction
Liu, What Does Context Compression Cost an Agent? Interaction Costs Unrevealed by Task-Completion Metrics The notebook keeps rules and diary apart; only the diary may be summarised
Liu et al., SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness Only huge tool outputs get trimmed, keeping the failures and the tail
Mendoza et al., Persistent LLM Orchestration of Heterogeneous Robot Teams via Partial Delegation to Multi-Agent Protocols recap rebuilds the lead's picture from the room's files, and the lead wakes on events

research

Paper What it changed
Lange et al., The Agentic Scientist: Orchestrating Coding Agents for Automated Scientific (Co-)Discovery Phases with gates, fixed roles, the sealed evaluator
Wang et al., Harness-Search: Guiding Long-Horizon Search through Multi-Agent Coordination The auditor and its verdicts at the gates
Zhang et al., PatchBoard: Schema-Grounded State Mutation for Reliable and Auditable LLM Multi-Agent Collaboration Agents propose; only the kernel and the registrar write the state
Li et al., MemTX: Transactional Belief Commit for Stateful Agent Memory Evidence stays tentative until it is curated, and a revoke withdraws what cited it
Tang et al., AI Research Agents Narrow Scientific Exploration Ideas alternate between gap and analogy, and near-duplicates are refused
Liu et al., IdeaScientist: Orchestrating Agents for Grounded Scientific Ideation Each idea comes as a full bundle, with its prior work
Virk et al., SwarmResearch: Orchestrating Coding Agents for Open-Ended Discovery One branch per idea, explorers and optimizers, the lead as a shepherd
Zhao et al., ScienceFlow: A long-horizon agent for ML research, scientific discovery and beyond A tag for every attempt and reanchor to restart from an old state
Pelleriti et al., What Do Evolutionary Coding Agents Evolve? Detectors for cycling and tuning-only runs
Cui et al., ARAC: Benchmarking Auto-Research's Alignment and Completeness on End-to-End Researchs The auditor's rubric, phase by phase
Ding et al., Autonomous Research Agents: A Survey of AI Scientists and the Verification Gap The report's verifiability checklist
Hu et al., The Winner's Curse in LLM Self-Improvement Loops: Selection Noise, Lock-in, and Acceptance Rules The best's margin in sigma and a run on fresh seeds in the report
Huang et al., RealisticTritonBench: A Benchmark for Triton-Kernel Generation in Real-World AI Frameworks The contract template for kernels: fresh inputs at each iteration
Feng et al., KREX: Concurrent Kernel Benchmarking on Shared GPUs via Region-Granular Exclusivity The same template: no side streams, noise thresholds
Prasad et al., BAITBENCH: Measuring Agent Reward Hacking with Optional Shortcuts Planted in ML Tasks A canary task with a planted shortcut, to test the room before trusting it
Yu et al., Beyond Execution: Auditing Experimental Fidelity in LLM-Driven Scientific Research A precheck on the data and the declared steps
Chen, When Does Combining Language Models Help? A Co-Failure Ceiling on Routing, Voting, and Mixture-of-Agents Across 67 Frontier Models research panel asks many free models and takes no vote
Fadnavis et al., Beyond Consensus: Trace-Level Synthesis in Mixture of Agents The panel keeps every answer whole, so an agent can read them all

The first loop of the research kernel, one file and a fixed evaluator with a commit when the number improves, comes from Andrej Karpathy's autoresearch.

Cite

@software{iancu_chatroom_2026,
  author = {Iancu, Adrian Ciprian},
  title  = {chatroom: a live room and a research kernel for coding agents},
  year   = {2026},
  url    = {https://github.com/Partido-De-What/claude-plugins}
}

Find me

GitHub · Hugging Face · LinkedIn · Instagram · site

License

AGPL-3.0-only. Versions up to 0.12.1 stay available under MIT; third-party credits are in NOTICE.

About

Claude Code and Codex plugins by ArtyomITA - chatroom: put your agents in one room and watch them work (live group chat, task board, 3D voxel office)

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages