Ross ROSS = Recommend OSS · open-source software intelligence for agents

sipyourdrink-ltd/bernstein

The open-source governance layer for AI agents: write the rules declaratively, Bernstein enforces them and produces the verifiable, replayable record. Free, Apache-2.0. https://bernstein.run observed · 2026-09-03

github.com/sipyourdrink-ltd/bernstein · homepage · Python · Apache-2.0 (permissive) observed · 2026-09-03

Health v2 · maintenance only

78/100

  • Activity 100
  • Release rhythm 88
  • Longevity 11

Flags: young

How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.

  • gap_med: 0.0
  • age_days: 164
  • days_rel: 2
  • days_push: 0
  • n_releases_24m: 151

Full methodology

Adoption not part of the score

1079 stars · 149 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Bernstein is a deterministic Python scheduler that orchestrates CLI coding agents (Claude Code, Codex, Gemini CLI, Aider, and 40+ more) running in parallel, each isolated in its own git worktree. It keeps the model out of the coordination loop so runs replay byte-identically, and provides signed lineage plus an opt-in HMAC audit chain that reviewers can verify offline.

Use cases

  • run multiple cli coding agents in parallel on one codebase
  • get deterministic, byte-identical replays of agent coding runs
  • isolate each agent task in its own git worktree
  • audit ai coding agent runs with a tamper-evident hmac log
  • deploy agent orchestration on-prem or air-gapped
  • gate agent-generated diffs on lint, types, and tests before merging
  • orchestrate claude code, codex, and gemini cli from one scheduler

When to choose

  • you run several coding agents in parallel and suffer nondeterministic merges
  • you need reproducible, provable agent runs with signed lineage for compliance
  • you want per-task git worktree isolation with quality gates before merge
  • you need air-gapped or on-prem agent orchestration
  • you want a model-free coordination loop so scheduling is deterministic

When to avoid

  • you need a single-agent assistant rather than a multi-agent orchestrator
  • you want the model itself to make coordination and routing decisions
  • you need a fully stable tool - it is beta software built by a solo maintainer
  • you prefer in-process plugin extensibility over subprocess-isolated adapters

Facets

cli-tool · maturity active

agent-framework workflow-automation developer-tools cli version-control security scheduling developer-tools large-language-models version-control security python cli self-hosted cross-platform agent-orchestrator deterministic-replay git-worktree audit-trail hmac-audit-log air-gap parallel-agents cli-coding-agents provenance reproducibility mcp-server multi-agent beta-software ai-agents automation linux macos docker

5 sources

Member repositories

RepositoryRoleHealth v2
sipyourdrink-ltd/bernsteinmain78

For agents

markdown · JSON · MCP: product_card(name="sipyourdrink-ltd/bernstein")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem