agent · ai agents frameworks

Bernstein - Multi-agent orchestration

bernstein.runevidence retrieved Aug 9, 2026 – Aug 18, 2026view as JSON all ai agents frameworks entries 29 alternatives

Indexed12 or more fields evidenced; not yet scored by a human against the rubric

21 fields evidenced · 38 with no public information. Every value below links to the document it came from and the date we checked it.

16 of those are quoted but not yet interpreted: we hold the source sentence and the date we read it, but nobody has written the answer to the question yet. Those rows show the quote and say so rather than repeating it back as an answer. Why these exist

2 of those are flagged for review: our own check found language in the cited source that may address the field. An editor has not adjudicated them yet.

1 absence has been read back against the source by an editor and confirmed.

3 of the evidenced values quote wording our own later capture of the same page no longer contains. Marked below, awaiting an editor.

Also appears in

Self-hosted and on-premise AI agents AI agents and vendors that state a tamper-evident audit trail

Trust gap

How this is scored

These scores are automated. A 1 or 2 counts how many fields this entry answers with a citation — not what those answers say. A 0 means every field for that dimension is recorded as undisclosed and none of the pages we captured mentions it. No one has scored this entry against the rubric yet.

Runtime governance

1/3 auto

Are scope, spend, and authority enforced while the agent runs?

Scored from 1 evidenced field(s): permission_scopes. Automated score; capped at 2 pending review.

Counts Permission scopes, each cited below.

Audit trail

2/3 auto

Is there a tamper-evident record of what it actually did?

Scored from 2 evidenced field(s): audit_trail, audit_trail_immutable. Automated score; capped at 2 pending review.

Counts Audit trail and Tamper-evident log, each cited below.

Compliance

unscored

Are compliance obligations attached per engagement?

Outcome settlement

unscored

Does payment depend on a verified result?

Insurance & indemnity

unscored

Can the deployment be insured, and is the customer indemnified?

Assurance

Insurance available: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Own liability cover: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Insurance carriers: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Coverage limits: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Indemnification: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Liability cap: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Audit trail: not yet interpreted
“flip on the hmac audit log and someone who did not run it can check the record offline, without rerunning it.”
bernstein.run · checked Aug 9, 2026
Tamper-evident log: not yet interpreted
“hmac-chained event log under .sdd/audit/ , one jsonl file per utc day. each entry references the previous hash. tampering breaks verification.”
bernstein.run · checked Aug 9, 2026
Explainability: not yet interpreted
“every step signed , in order, on disk. run with --audit and bernstein writes an hmac-chained event log under .sdd/audit/ , one jsonl file per utc day. each entry references the previous hash.”
bernstein.run · checked Aug 9, 2026
Runtime governance: no public information
bernstein.run · checked Aug 18, 2026 · source did not state this
Permission scopes: not yet interpreted
“isolates each in its own git worktree”
bernstein.run · checked Aug 9, 2026
Compliance certifications: no public information
bernstein.run · checked Aug 9, 2026 · absence verified against this source
Regulatory alignment: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Outcome-based pricing: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Settlement mechanism: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Asset custody: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Dispute process: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
SLA terms: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Evaluation coverage: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Self-reported performance: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this

Agency

Autonomy level: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Human oversight: No model sits in the coordination loop, so a run replays byte-identically and can be checked by a reviewer who did not execute it. The earlier "always-on replay journal" wording is no longer on the page.
“no model in the coordination loop, so the same plan replays byte-identically”
bernstein.run · checked Aug 18, 2026
Goal complexity: Decomposes a goal into tasks and runs CLI coding agents in parallel, each in its own git worktree, with lint, types and tests gating every merge. Adapter count is now stated as "40+ more" rather than the earlier "42 other agents".
“bernstein is a deterministic python scheduler that runs cli coding agents in parallel.”
Not in our latest capture. This quote was on the page when we filed it. We captured the same page again on Sep 17, 2026 and it was not there. The value above has not been re-checked by an editor yet. How we re-check citations
bernstein.run · checked Aug 18, 2026
Action space: not yet interpreted
“runs cli coding agents in parallel isolates each in its own git worktree verifies lint · types · tests per diff merges only what passes”
bernstein.run · checked Aug 9, 2026
Operating environment: not yet interpreted
“runs on your laptop, on-prem, air-gapped”
bernstein.run · checked Aug 9, 2026
Initiative: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this

Safety

Safety evaluations: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Red teaming: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Safety policy: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Usage restrictions: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Model or system card: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Incident reporting: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Third-party evaluations: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Data handling: not yet interpreted
“nothing leaves your machine. someone who did not run it can check it. signature and hash-chain checks read the files alone; the hmac leg needs the key the chain was written with, so hand a reviewer both and bernstein audit verify answers without rerunning anything.”
bernstein.run · checked Aug 9, 2026

Practicality

Pricing model: not yet interpreted
“~$0 to install”
bernstein.run · checked Aug 9, 2026
Price point: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Availability: not yet interpreted
“open-source”
bernstein.run · checked Aug 9, 2026
Deployment options: no public information
Flagged for review — the source below contains language that may address this field. Not yet checked by an editor. How we check absences
bernstein.run · checked Aug 9, 2026
Integrations: not yet interpreted
“ships adapters for claude code, codex, gemini cli, aider, and 40+ more”
bernstein.run · checked Aug 9, 2026
Supported regions: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Support model: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this

Foundation models

Base models: Orchestrates Claude Code, Codex, Gemini CLI and other CLI coding agents in parallel git worktrees
“Orchestrate Claude Code, Codex, Gemini CLI and other CLI coding agents in parallel git worktrees.”
bernstein.run · checked Aug 18, 2026
Model provider: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Model swappable: Model selection is LLM-assisted via a capability router and recommender, with best-of-n judged by an LLM; both opt-in and pluggable.
“model selection is llm-assisted (capability router + recommender), and best-of-n picks a winner via llm judge”
bernstein.run · checked Aug 18, 2026
Open weights: not yet interpreted
“apache 2.0”
bernstein.run · checked Aug 9, 2026
Fine-tuning: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Context window: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this

Ecosystem

Protocols supported: no public information
bernstein.run · checked Aug 18, 2026 · source did not state this
Tool use: not yet interpreted
“per-task tool calls”
bernstein.run · checked Aug 9, 2026
Multi-agent: not yet interpreted
“multi-agent orchestration”
bernstein.run · checked Aug 9, 2026
API access: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Open source: no public information
Flagged for review — the source below contains language that may address this field. Not yet checked by an editor. How we check absences
bernstein.run · checked Aug 9, 2026
Marketplace presence: not yet interpreted
“picked up by awesome lists, newsletters & peer projects”
bernstein.run · checked Aug 9, 2026

Impact

User base: 72 contributors
“72 contributors ~$0 to install · sponsor → /cost”
Not in our latest capture. This quote was on the page when we filed it. We captured the same page again on Sep 17, 2026 and it was not there. The value above has not been re-checked by an editor yet. How we re-check citations
bernstein.run · checked Aug 18, 2026
Deployment scale: not yet interpreted
“2,463 closed prs”
Not in our latest capture. This quote was on the page when we filed it. We captured the same page again on Sep 17, 2026 and it was not there. The value above has not been re-checked by an editor yet. How we re-check citations
bernstein.run · checked Aug 18, 2026
Target sectors: not yet interpreted
“CLI coding agents”
bernstein.run · checked Aug 9, 2026
High-risk domains: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Documented incidents: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this
Market recognition: no public information
bernstein.run · checked Aug 9, 2026 · source did not state this