Releases·Tool Release
Running several Claude Code agents on one repo: branches hid the real conflict in all 5 runs, a shared directory fixed it, and the agents started messaging each other on their own
A developer released Médula, an MIT-licensed kernel that coordinates multiple Claude Code agents (Sonnet 5, effort high) working on one repo. In tests with a room-booking API and 37 acceptance tests, a shared working directory scored 37/37 in all 10 runs versus 31/37 with git branches. Unprompted agent-to-agent messaging also emerged as a possible alternative coordination mechanism.
6 engagement1 sourceThu, Oct 1, 2026, 12:17 PM
AI Overview
- Author's open-source MIT experiment (Médula, on GitHub) ran 6 colliding-by-meaning tasks and 37 acceptance tests on a room-booking API with Sonnet 5 agents.
- Sequential mode scored 30/37; the export agent refused to skip the new OTP second factor but awaited a confirmation that never arrives non-interactively.
- One-branch-per-task merges scored 31/37 across all 5 runs: agents resolved textual conflicts while /export kept calling login() without the OTP.
- A shared working directory scored 37/37 in all 10 runs, with per-file locks or the Médula kernel (Claude Code hooks PreToolUse/PostToolUse).
- Médula matched locks' $1.65 per run but caught 6/6 conflicts vs 5 and made 0 unnecessary blocks vs 5; locks once stalled an agent 453 s.
- With messaging tools enabled, agents communicated unprompted — in one run the renaming agent warned the filtering agent; tools now disabled for comparability.
- Caveats: 1 to 5 runs per mode, and calibration labels were written by Claude.
Related
Releases·Tool Release
Open-source test of AI coding agents on one repo: isolated branches broke in all 5 runs, shared dirs and agent messaging passed
A developer launched Médula, a free MIT-licensed tool that asks a fast decision model before every write whether a change collides with another agent's work. His experiment found that six agents on separate branches each passed their own tests while the merged app broke in all 5 runs, while agents sharing one working directory passed in all 10 runs. The finding matters to anyone running multi-agent coding workflows: text merges hide semantic conflicts, and letting agents see or message each other fixes them.
A developer launched Médula, a free MIT-licensed tool that asks a fast decision model before every write whether a change collides with another agent's work. His experiment found that six agents on separate branches each passed their own tests while the merged app broke in all 5 runs, while agents sharing one working directory passed in all 10 runs. The finding matters to anyone running multi-agent coding workflows: text merges hide semantic conflicts, and letting agents see or message each other fixes them.
41 engagement2 sourcesredditFri, Oct 2, 2026, 07:15 AM
Products·building
Médula: MIT-licensed kernel coordinating parallel Claude Code agents on one repo, with published decision logs
Médula is an open-source lab and kernel that lets multiple Claude Code agents work on the same repo in parallel, using a cheap decision model to arbitrate conflicts. It ships as a pluggable-decider experiment with a small API, acceptance tests, and fully public run data (sessions, diffs, SQLite decision logs). It targets developers running multi-agent coding setups who need agents to avoid breaking each other's changes.
Médula is an open-source lab and kernel that lets multiple Claude Code agents work on the same repo in parallel, using a cheap decision model to arbitrate conflicts. It ships as a pluggable-decider experiment with a small API, acceptance tests, and fully public run data (sessions, diffs, SQLite decision logs). It targets developers running multi-agent coding setups who need agents to avoid breaking each other's changes.
13 engagement1 sourceredditFri, Oct 2, 2026, 07:10 AM
Products·launched
Médula: kernel coordinating parallel Claude Code agents via a fast LLM collision decider
Médula is an MIT-licensed lab plus kernel that lets multiple Claude Code agents work on one repo concurrently. A PreToolUse hook sends each write/edit/command intent to a fast LLM decider (Jev) that returns a collision probability, gating whether agents proceed or wait. It targets developers running parallel coding agents on a shared codebase.
Médula is an MIT-licensed lab plus kernel that lets multiple Claude Code agents work on one repo concurrently. A PreToolUse hook sends each write/edit/command intent to a fast LLM decider (Jev) that returns a collision probability, gating whether agents proceed or wait. It targets developers running parallel coding agents on a shared codebase.
11 engagement1 sourceredditThu, Oct 1, 2026, 12:14 PM
Social·Community
Every Agent Passed Its Tests. The Merge Was Broken.
A Medium post describes running several Claude Code agents on a single repository: each agent passed its own tests, yet Git merges still let real conflicts through. The author reports the fix was not at the kernel level, highlighting merge integration as a weak point in multi-agent coding workflows.
A Medium post describes running several Claude Code agents on a single repository: each agent passed its own tests, yet Git merges still let real conflicts through. The author reports the fix was not at the kernel level, highlighting merge integration as a weak point in multi-agent coding workflows.
0 engagement2 sourcesrssFri, Oct 2, 2026, 10:37 AM
Social·Community
Developer reports two-agent workflow failure after Codex and Claude Code collide on shared repo
A developer tried running Claude Code on a backend task and Codex on a frontend task in the same repository, which worked for about two hours before Codex refactored a types file Claude was actively using. Neither agent detected the conflict, leading to a merge that required most of the next morning to untangle.
A developer tried running Claude Code on a backend task and Codex on a frontend task in the same repository, which worked for about two hours before Codex refactored a types file Claude was actively using. Neither agent detected the conflict, leading to a merge that required most of the next morning to untangle.
10 engagement1 sourceredditWed, Aug 12, 2026, 01:19 PM