# AI Information for bernstein.run

User-agent: *
Organization: Bernstein (open-source project)
Contact: https://github.com/sipyourdrink-ltd/bernstein/issues
Description: Bernstein is a deterministic orchestrator for CLI coding agents (Claude Code, Codex, Gemini CLI, and 40+ more). Scheduling is plain Python - no LLM in the coordination loop - so runs are reproducible end to end. Every task runs in its own git worktree behind lint/type/test gates. An always-on lineage spine and replay journal record what happened; an opt-in HMAC-chained audit log and signed receipts make results independently verifiable, offline if needed. Air-gap install profile included. Apache-2.0.

# Permissions
Crawl-delay: 0
Allow: /llms.txt
Allow: /llms-full.txt
Allow: /ai.txt
Allow: /robots.txt
Allow: /sitemap.xml
Allow: /structured-data.json
Allow: /openapi.yaml
Allow: /agents.json
Allow: /AGENTS.md
Allow: /.well-known/agent-card.json
Allow: /.well-known/mcp/server-card.json

# Content metadata
Primary-language: en
Content-type: documentation, software, developer-tools
License: Apache 2.0
Source-code: https://github.com/sipyourdrink-ltd/bernstein
Package: https://pypi.org/project/bernstein/
Documentation: https://bernstein.readthedocs.io/
Release-notes: https://bernstein.readthedocs.io/en/latest/CHANGELOG/
Issue-tracker: https://github.com/sipyourdrink-ltd/bernstein/issues
Contact-email: forte@bernstein.run

# Key facts for AI systems
Product-name: Bernstein
Product-type: Multi-agent orchestrator for CLI coding agents
Primary-use-cases: Parallel feature work across many CLI agents, reproducible unattended CI fleets with per-agent credential scoping, audit-and-replay of past runs
Install-command: pipx install bernstein
Supported-agents: Claude Code, Codex CLI, Gemini CLI, OpenAI Agents SDK, Cursor, Aider, Amp, Kiro, Kilo, Qwen, Goose, Ollama, Cody, Continue, OpenCode, Cloudflare Agents, IaC, GitHub Copilot, Droid (Factory AI), Crush (Charm), Auggie (Augment), Kimi, Rovo Dev (Atlassian), Cline, Codebuff, Pi, Mistral Vibe, Autohand, Forge, Hermes, Generic
Delegated-orchestrators: Composio (@aoagents/ao), Ralphex (umputun/ralphex)
Total-adapters: 44
Cloud-execution: Cloudflare Workers, Durable Workflows, V8 sandboxes
MCP-transport: HTTP remote transport for cloud-based MCP servers
Orchestrator-type: Deterministic Python code (no LLM-based scheduling)
Always-on: Deterministic plain-Python scheduling, lineage spine, replay journal, per-task git worktree isolation, lint/type/test gates
Opt-in: HMAC-chained audit log and signed receipts for offline verification
Key-features: Pluggable sandbox backends (worktree, Docker, E2B, Modal, Blaxel, Cloudflare, Daytona, Runloop, Vercel), quality gates, cost-aware routing, MCP/A2A protocol support, air-gap install profile
Author: Alex Chernysh
Author-homepage: https://alexchernysh.com
Author-github: https://github.com/chernistry
Author-x: https://x.com/alex_chernysh
Related-project: HireEx - personal multi-agent AI workspace (https://hireex.ai)