| What it does | The verification layer for AI code change
| Air traffic control for your coding agents
|
|---|
| Stage | MVP
| Startup
|
|---|
| Founded | Not given
| Not given
|
|---|
| Based in | Not given
| Not given
|
|---|
| Pricing model | Freemium
| Open source
|
|---|
| Pricing | Free plan: $0/mo with 500 verifications/mo and unlimited local CLI runs. Pro: $59/mo with 10,000 verifications/mo (early access). Scale: $300/mo with 100,000 verifications/mo (early access). Enterprise: Custom pricing (early access).
| Not given
|
|---|
| Key features | - Verify arbitrary diffs from agents, codemods, or humans
- Three-verdict system: SAFE, UNSAFE, UNPROVEN
- Runs test suite on isolated shadow tree copy
- Changed-line coverage attestation for Python
- Reproducible verification report with audit-grade record
- MCP server for AI agents to verify before landing
- CI gate that blocks unsafe merges
- Syntax and imports validation gates
- Runs locally with no code upload to external service
| - Task queuing with dibs run wrapping commands
- Real-time job status board for all agents
- Memory and disk resource monitoring
- Agent-to-agent messaging with dibs tell and dibs say
- Human approval workflow for irreversible actions
- Browser access via end-to-end encrypted relay at app.dibs.cc
- MCP client support for Claude with read-only or action modes
- Multi-machine job distribution with --anywhere flag
- Job scheduling at specific times with dibs later
- API rate limit sharing with quota buckets
|
|---|
| What makes it different | - Only system that proves changed lines are actually tested, not just that tests pass
- Works with any AI coding agent and integrates via MCP
- Code never leaves your machine; runs entirely locally
- Free local CLI with unlimited verifications, paid only for managed infrastructure
| - Prevents agent collisions with deterministic turn-taking on shared resources
- Automatically manages machine memory and disk to keep builds running
- Agents communicate peer-to-peer while humans only handle critical decisions
- Fully open source with no per-seat or subscription fees
- Works across multiple machines with a single encrypted board
|
|---|
| Community votes | 0 | 0 |
|---|
| DR (Domain Rating by Ahrefs) | 2 ● no change since last check | Not checked yet ● no change since last check |
|---|
| Trust Flow (Majestic) | 0 | 0 |
|---|
| Citation Flow (Majestic) | 15 | 0 |
|---|
| Referring domains (Majestic) | 4 | 0 |
|---|
Refactron™