naw103/foremerge

Catch intent conflicts before code conflicts. The open-source coordination protocol for coding agents, built above Git.

What it solves

Foremerge addresses "intent conflicts" that occur when multiple AI coding agents work on the same project simultaneously. While Git can detect when two agents edit the same line of code, it cannot detect when two logically sound changes in different files contradict each other (e.g., one agent replaces a service that another agent is currently extending). Foremerge allows agents to coordinate their plans before writing any code, preventing wasted effort and broken integrations.

How it works

Foremerge acts as a coordination protocol built on top of Git. It uses a shared SQLite database stored within the project's .git folder, allowing agents operating in separate Git worktrees to share a "semantic whiteboard."

  1. Intent Publication: Before coding, an agent announces its target (e.g., "I am changing the sendEmail function") and the operation (e.g., replace or extend).
  2. Conflict Detection: Foremerge compares these declarations against other agents' plans. If two agents target the same symbol with conflicting operations, it raises an advisory warning.
  3. Verification Gate: Once work is complete, Foremerge requires a validation check (like a test suite) to run against the specific code fingerprint before the work can be marked as accepted.
  4. Non-Blocking: It does not lock files or block agents; it provides advisory warnings and evidence so the human or agent can decide how to proceed.

Who it’s for

Developers using multiple AI coding agents (such as Claude Code, Codex, or Cursor) in parallel on a single codebase who need to prevent logical collisions and maintain an audit trail of agent decisions.

Highlights

  • Semantic Awareness: Detects conflicts based on symbols and intents rather than just text lines.
  • Local-First: Uses a local SQLite store in the .git directory for zero-config sharing between local worktrees.
  • MCP Integration: Provides a Model Context Protocol (MCP) server to allow AI agents to natively register, publish intents, and query conflicts.
  • Verification-Gated: Ensures that ChangeSets are validated by actual command execution (e.g., cargo test) rather than relying on agent-reported status.

Related

  • Dispatch
    Foremerge 0.5.0 – Intent‑Based Conflict Detection for Parallel Coding Agents

    Foremerge 0.5.0 introduces a local‑first protocol that lets AI coding agents publish their intended semantic changes before editing, allowing deterministic, advisory conflict warnings that Git alone cannot provide.

  • Project
    spacering-net/codeg

    A multi-agent coding workspace that integrates various AI coding agents into a single visual interface, allowing them to collaborate and work on the same project.

  • Project
    unclebob/swarm-forge

    A coordination framework that manages AI agents in isolated git worktrees and tmux sessions, using a durable handoff protocol to exchange work.

  • Project
    peterjthomson/ledger

    A modernised git interface for macOS that helps developers oversee and manage code changes generated by AI coding agents and software teams.

  • Project
    AVIDS2/memorix

    Memorix is a Node‑JS CLI/library that adds a local, SQLite‑backed, searchable memory store to any Git project. It lets AI coding assistants (Claude Code, Copilot, Cursor, etc.) share project facts, code snapshots, design rationales and commit‑derived knowledge across sessions and tools via the Model‑Context‑Protocol. Features include autopilot worksets, observation/reasoning memory, Git‑memory ingestion, code‑graph indexing, orchestration of multi‑agent tasks, a web dashboard, and optional sync. Install with npm, run `memorix setup --agent <agent>` and the agents can start reusing the shared memory.