Vigla

Open-source mission control for AI coding agents

Free· Free and open source under Apache-2.0. No paid tiers; users build from source.AgentsModel-agnostic (delegates to the underlying agent's LLM: Claude, GPT, etc.)

Visit website →

Best for

Individual engineers or small dev teams on macOS who run multiple AI coding agents in parallel on the same repo and want an audited, worktree-isolated review gate before anything hits main.

Skip if

Windows or Linux developers, non-technical users who want a click-to-install product, or teams that need a hosted, multi-user control plane with SSO and shared state.

Vigla is an open-source mission-control desktop app for supervising AI coding agents. Instead of replacing agents like Claude Code, Codex CLI, or Antigravity, it wraps them in a governance layer that runs each worker in an isolated git worktree, audits the resulting diff against scope, tests, and risk criteria, and only then lets you merge the change back into your main branch. The design goal is to sit between agent output and your protected branch, so teams get parallelism and delegation without the usual blast radius of an autonomous coder pushing directly to a shared tree.

Each 'mission' spawns a worktree per agent, records the prompt, tool calls, and file changes, and produces an auditable trail that can be reviewed, replayed, or rolled back. The pre-merge audit is opinionated: it flags out-of-scope edits, missing or failing tests, and risky patterns before you ever look at the diff yourself, which turns multi-agent orchestration from a trust exercise into a review workflow.

Because the control plane is local Rust with no vendor lock-in and no telemetry, Vigla is a natural fit for engineers who want to run several coding agents in parallel on the same repo but keep review, secrets, and merge policy under their own laptop. It is currently macOS 12+ only and ships as source you build yourself with Rust 1.95, Node 22, pnpm 10, and Xcode command line tools, so it targets developers comfortable with a toolchain build rather than click-to-install users.

Editor's take

Vigla is a sharp, honest take on the missing piece around today's coding agents: not another agent, but the review and rollback plane that lets you actually let them work. The macOS-only, build-from-source posture keeps it firmly in early-adopter territory, but the worktree isolation plus pre-merge audit is exactly the shape I want when running Claude Code and Codex side by side.

— The AI Tool Bible editorial team

Pros

✅ Cross-vendor: coordinates Claude Code, Codex CLI, Antigravity, and other coding agents from one cockpit
✅ Isolated git worktrees per agent prevent parallel workers from stepping on each other
✅ Pre-merge audit checks scope, tests, and risk before changes reach your main branch
✅ Full mission audit trail with reversible rollback for anything that ships
✅ Local control plane, no telemetry, no cloud dependency, Apache-2.0 licensed
✅ Model-agnostic — you keep using whatever LLM your chosen agent already ships with

Cons

⚠️ macOS 12+ only today; no Linux or Windows builds
⚠️ No pre-built binaries — requires a Rust/Node/pnpm/Xcode toolchain build from source
⚠️ Early-stage project with a small footprint and limited documentation compared to established agent frameworks
⚠️ No hosted or team collaboration layer; state lives on the operator's laptop
⚠️ No external API for programmatic control from other systems

Use cases

Supervising multiple AI coding agents in parallelPre-merge audit of agent-generated diffsIsolating agent work in per-mission git worktreesRolling back an agent mission that went wrongComparing Claude Code vs Codex CLI on the same taskKeeping AI coding governance fully local with no telemetryAuditable trail of what an agent changed and why