Content
<div align="center">
<picture>
<source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/sipyourdrink-ltd/bernstein/main/docs/assets/logo-dark.svg">
<source media="(prefers-color-scheme: light)" srcset="https://raw.githubusercontent.com/sipyourdrink-ltd/bernstein/main/docs/assets/logo-light.svg">
<img alt="Bernstein" src="https://raw.githubusercontent.com/sipyourdrink-ltd/bernstein/main/docs/assets/logo-light.svg" width="340">
</picture>
<br>
<img alt="Bernstein - deterministic multi-agent CLI orchestration" src="https://raw.githubusercontent.com/sipyourdrink-ltd/bernstein/main/docs/assets/banner-readme.png" width="820">
<br>
> *"To achieve great things, two things are needed: a plan and not quite enough time."* - [attributed to](https://quoteinvestigator.com/2020/08/19/plan-time/) Leonard Bernstein
### deterministic multi-agent CLI orchestration
[](https://github.com/sipyourdrink-ltd/bernstein/actions/workflows/ci.yml)
[](https://pypi.org/project/bernstein/)
[](https://ghcr.io/sipyourdrink-ltd/bernstein)
[](https://python.org)
[](https://github.com/sipyourdrink-ltd/bernstein/blob/main/LICENSE)
[](https://scorecard.dev/viewer/?uri=github.com/sipyourdrink-ltd/bernstein)
[](https://github.com/sipyourdrink-ltd/bernstein/actions/workflows/codeql.yml)
[](https://codespaces.new/sipyourdrink-ltd/bernstein?quickstart=1)
[](https://mcptoplist.com/server/io.github.sipyourdrink-ltd%2Fbernstein)
<a href="https://deepwiki.com/sipyourdrink-ltd/bernstein"><img src="https://deepwiki.com/badge.svg" alt="Ask DeepWiki"></a>
[website](https://bernstein.run) · [docs](https://bernstein.readthedocs.io/) · [install](https://bernstein.readthedocs.io/en/latest/getting-started/install/) · [first run](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/getting-started/first-run.md) · [glossary](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/reference/GLOSSARY.md) · [limitations](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/reference/KNOWN_LIMITATIONS.md) · [name policy](https://github.com/sipyourdrink-ltd/bernstein/blob/main/TRADEMARKS.md) · [sponsor](https://github.com/sponsors/chernistry)
[简体中文](https://github.com/sipyourdrink-ltd/bernstein/blob/main/README.zh-Hans.md) · [繁體中文](https://github.com/sipyourdrink-ltd/bernstein/blob/main/README.zh-TW.md)
</div>
---
> **Status: beta.** Solo-maintained, under active development. The version number counts releases, not maturity - minor versions may change interfaces. Pin the version for anything you depend on; regressions get fixed fast, [file them](https://github.com/sipyourdrink-ltd/bernstein/issues).
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 coding task runs in its own git worktree behind lint/type/test gates; artifact-mode tasks, which complete on a signed lineage receipt instead of a commit, get a plain working directory instead. Results stay checkable after the fact: an always-on lineage spine and replay journal, plus an opt-in HMAC-chained audit log (`BERNSTEIN_AUDIT=1`) with receipts you can verify offline. Air-gap install profile included. Apache-2.0.
### at a glance
Four things set it apart; everything after is detail.
- **No LLM in the coordination loop.** Scheduling is plain Python, so a run is reproducible end to end. Replay yesterday's plan and get yesterday's task graph.
- **Checkable after the fact.** The replay journal records every run, and the always-on lineage spine records every lineage-bearing step; the opt-in audit chain adds receipts you verify offline. Non-determinism surfaces as a hash mismatch at the exact step, not a flaky re-run. Non-code deliverables get the same treatment: a task can declare an artifact contract (report, dataset, action log, ops result) on a plan step, a backlog entry, or the task CLI and complete on a signed lineage receipt instead of a git commit.
- **Isolated by construction.** Each coding task gets its own git worktree behind merge gates; artifact-mode tasks get working-directory separation under `.sdd/workspaces/`. Under this default isolation agents share no mutable workspace; coordination state (the task backlog) is shared and claimed atomically. Filesystem enforcement beyond that separation is opt-in, from the [sandbox backends](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/architecture/sandbox.md) (disabling worktrees runs every task in the shared checkout).
- **Broad and local.** 40+ CLI agent adapters plus a generic `--prompt` wrapper, file-based state, no SaaS hop, no third-party data plane.
The full list is on the [capabilities page](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/reference/capabilities.md); the [feature matrix](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/reference/FEATURE_MATRIX.md) is the exhaustive index.
### install in 30 seconds
```bash
uv tool install bernstein # or: pipx install bernstein
bernstein init
bernstein doctor # checks a CLI agent is installed and authenticated
bernstein -g "fix the failing test in tests/test_foo.py"
```
pipx, pip, brew, dnf, npm, and Docker are covered in the [install guide](https://bernstein.readthedocs.io/en/latest/getting-started/install/); the air-gapped wheelhouse has its own [air-gap guide](https://bernstein.readthedocs.io/en/latest/installation/air-gap/).
<img alt="A real bernstein demo run: mock agents fix four seeded bugs, ending on the run's signed receipt verifying offline" src="https://raw.githubusercontent.com/sipyourdrink-ltd/bernstein/main/docs/assets/demo-run/demo.gif" width="820">
The recording above is a real run, and it ships with its own proof: the cast, the signed run receipt derived from that run's journal, and the public key that pins it live together in [`docs/assets/demo-run/`](https://github.com/sipyourdrink-ltd/bernstein/tree/main/docs/assets/demo-run). Verify the run you just watched, offline:
```bash
bernstein verify receipt docs/assets/demo-run/run-receipt.json \
--public-key docs/assets/demo-run/run-receipt.pub.pem
```
CI re-verifies the committed receipt on every push to main — and proves a tampered copy fails — so the published evidence cannot rot into a decorative file. `scripts/record_demo.sh` regenerates the recording, receipt, and key from a fresh real run; nothing inside the terminal is synthesised.
A run in flight is watchable from either operator surface. Both read the same task API, so neither is a lagging mirror of the other.
|  |  |
|:---:|:---:|
| `bernstein live` — the terminal dashboard | `bernstein gui serve` — the browser dashboard |
### prove a run
Determinism here is something you check, not something you take on faith. Run once with audit enabled, then verify what was recorded:
```bash
BERNSTEIN_AUDIT=1 bernstein -g "fix the failing test in tests/test_foo.py"
bernstein replay list # run ids recorded on disk
bernstein replay latest --verify # recompute the journal head, name the first divergent step
bernstein lineage verify <run_id> # recompute the always-on lineage spine
bernstein audit verify # HMAC chain + Merkle seal (written because audit was enabled)
bernstein audit diagnose <run_id> --signal gate --sign-key KEY
# name the exact step a failure entered the run, as a signed receipt
bernstein verify run <run_id> --signing-key-path key.pem # sign one portable run receipt
bernstein verify receipt .sdd/runs/<run_id>/run-receipt.json # verify it offline: file only
```
The journal is written on every run; the lineage spine is always on and gains an entry for each lineage-bearing step, so a short run can finish with a valid, empty spine. `bernstein audit verify` only has a chain to check when the run was started with `BERNSTEIN_AUDIT=1`, a compliance preset, or `bernstein run --audit`. The `--audit` flag belongs to `bernstein run`; on the `bernstein -g` form above, set the environment variable.
The run receipt binds the journal head - and the lineage-spine head when the run wrote spine entries - plus, opt-in, an audit-chain range, under one Ed25519-signed subject with the public key embedded, so a reviewer holding the file and the operator's public key can confirm the recorded actions are exactly what executed - no HMAC key, no live `.sdd/`, exit `2` naming the first divergent step on tamper. With the file alone (no `--public-key` pin) the check is integrity-only: it proves the receipt is internally consistent, not who signed it, and the verdict says so. Details in [deterministic replay](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/operations/deterministic-replay.md#signed-run-receipt-one-file-offline-verification).
The same checkability applies to evaluation numbers: `bernstein bench run <suite> --reliability k` (also spelled `bernstein eval --reliability k`) runs every task `k` times under fixed coordination and reports a `pass^k` floor (all `k` attempts must pass) alongside the `pass@1` ceiling, sealed in a signed receipt that `bernstein bench reliability-verify` recomputes offline — a fabricated floor fails verification. Details: [pass^k reliability floor](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/eval/reliability.md).
### how it works
Each goal moves through four stages:
1. **Decompose**. The manager breaks your goal into tasks with roles, owned files, and completion signals. One LLM call, then plain Python from there.
2. **Spawn**. Agents start in isolated [git worktrees](https://git-scm.com/docs/git-worktree), one per coding task; an artifact-mode task gets a plain working directory instead. Main branch stays clean.
3. **Verify**. The janitor checks concrete signals: tests pass, files exist, lint clean, types correct.
4. **Merge**. Verified work lands in main. Failed tasks get retried or routed to a different model.
Why the scheduler is plain Python, and what that trades away: [why deterministic](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/architecture/WHY_DETERMINISTIC.md).
### everyday commands
```bash
cd your-project
bernstein init # creates .sdd/ workspace, bernstein.yaml + templates/
bernstein -g "Add rate limiting" # agents spawn, work in parallel, verify, exit
bernstein live # watch progress in the TUI dashboard
bernstein run plan.yaml # multi-stage plan: skip LLM planning, execute directly
bernstein stop # graceful shutdown with drain
```
The full operator surface (PR automation, schedules, chat bridges, the autofix daemon) is in [operator commands](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/operations/commands.md).
Repository hygiene gates: `bernstein readme-l10n verify` fails a PR whose translated READMEs drifted from the English source (naming the stale section), `bernstein readme-l10n sync` rebinds them after an English edit. See [readme-l10n](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/playbooks/readme-l10n.md).
### supported agents
Claude Code, Codex CLI, Gemini CLI, GitHub Copilot CLI, Cursor, Aider, Goose, Muse Code, OpenAI Agents SDK, Amp, Cody, Continue, Devin Terminal, Junie, Kilo, Kiro, AWS Q Developer, Ollama, OpenCode, OpenHands, Open Interpreter, gptme, Plandex, AIChat, Letta Code, Qwen, and more. The [adapter index](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/adapters/index.md) carries install commands for 30 of them; `bernstein integrations list` enumerates all 51 wired-in integrations from the registry in `src/bernstein/adapters/registry.py`, which is the single source of truth for what resolves - 49 of them are selectable agent adapters, the other two rows being the `mock` test stub and the `self-hosted-endpoints` endpoint profile; `src/bernstein/adapters/use_cases.py` carries the end-user copy for each one. Anything else with a `--prompt` flag works through the generic wrapper.
Mix agents in the same run: cheap local models for boilerplate, heavier cloud models for architecture. `bernstein integrations list --installed` shows what is available on your machine.
### beyond the front page
Everything deep lives on the [docs site](https://bernstein.readthedocs.io/):
| page | what it covers |
|---|---|
| [capabilities](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/reference/capabilities.md) | the full capability list: MCP server mode, signed agent cards, sandbox backends, artifact sinks, regulatory mappings |
| [who this is for](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/use-cases.md) | where the value lands, and where Bernstein is the wrong tool |
| [workflows](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/operations/workflow-manifests.md) | declarative YAML DAGs of agent / command / loop nodes |
| [web UI](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/gui/index.md) | browser dashboard on the same API the TUI uses |
| [cloud execution](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/cloudflare/cloudflare-overview.md) | experimental: run agents on Cloudflare Workers with R2 workspace sync against your own account. The hosted `api.bernstein.run` service is not yet available |
| [datasources](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/operations/datasources.md) | read-only query receipts, plus a query driver that binds each result to the schema snapshot it was derived against |
| [security](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/operations/security.md) | scorecard, fuzzing, hardening |
| [architecture](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/architecture/ARCHITECTURE.md) | how it works under the hood |
### why the name?
Bernstein is named after Leonard Bernstein, the American conductor and composer. The project orchestrates a crew of CLI coding agents the way Bernstein conducted the New York Philharmonic: every player on cue, the score deterministic, the conductor accountable for the result.
i wrote bernstein because i was paying $400/month in claude bills running three coding agents in parallel and getting nondeterministic merges. Apache 2.0, solo maintained. Live stats: [bernstein.run](https://bernstein.run).
### mentioned in
Listed in [vinta/awesome-python](https://github.com/vinta/awesome-python), covered in Augment Code's [open-source agent orchestrators](https://www.augmentcode.com/tools/open-source-agent-orchestrators) roundup, and listed in [Python Weekly #742](https://www.pythonweekly.com/p/python-weekly-issue-742-april-23-2026). We also wrote up the approach as the [deterministic zero-LLM orchestration](https://github.com/nibzard/awesome-agentic-patterns/blob/main/patterns/deterministic-zero-llm-orchestration.md) pattern in awesome-agentic-patterns.
<details>
<summary>All coverage: 20+ awesome lists, directories, newsletters, and peer citations</summary>
<br>
The full tracked list, including every awesome-list entry, catalog listing, prior-art citation, and newsletter mention, lives in [docs/mentions.md](https://github.com/sipyourdrink-ltd/bernstein/blob/main/docs/mentions.md). Entries are added as they appear; corrections welcome by issue or PR.
</details>
### contributing, support, license
PRs welcome; [CONTRIBUTING.md](https://github.com/sipyourdrink-ltd/bernstein/blob/main/CONTRIBUTING.md) has setup and code style. Security reports go through [SECURITY.md](https://github.com/sipyourdrink-ltd/bernstein/blob/main/SECURITY.md). If Bernstein saves you time: [GitHub Sponsors](https://github.com/sponsors/chernistry). Contact: [forte@bernstein.run](mailto:forte@bernstein.run).
Citation metadata lives in [CITATION.cff](https://github.com/sipyourdrink-ltd/bernstein/blob/main/CITATION.cff). License: [Apache-2.0](https://github.com/sipyourdrink-ltd/bernstein/blob/main/LICENSE); the project name is covered separately in [TRADEMARKS.md](https://github.com/sipyourdrink-ltd/bernstein/blob/main/TRADEMARKS.md).
---
[Alex Chernysh](https://alexchernysh.com) · [GitHub](https://github.com/chernistry) · [X](https://x.com/alex_chernysh) · [bernstein.run](https://bernstein.run)
<!-- mcp-name: io.github.sipyourdrink-ltd/bernstein -->
