Question 1

What is Molecule AI, in one sentence?

Accepted Answer

Molecule AI is the open-core control plane that lets heterogeneous AI agents — LangGraph, Claude Code, CrewAI, AutoGen, DeepAgents, OpenClaw — run as governed, memory-scoped teams instead of disconnected demos.

Question 2

How is it different from agent frameworks like LangGraph or CrewAI?

Accepted Answer

Those frameworks solve the execution layer — how an agent runs, calls tools, and hands off. Molecule AI solves the organizing layer above them: durable workspace roles, hierarchy-aware memory, governance, and operations. Many agent runtimes, one organizational operating system.

Question 3

Which runtimes are supported today?

Accepted Answer

Six adapters are shipping on the main branch of the monorepo: LangGraph, DeepAgents, Claude Code, CrewAI, AutoGen, and OpenClaw. NemoClaw (NVIDIA-oriented) is work in progress on a feature branch and will land in main once merged.

Question 4

Is it really open source?

Accepted Answer

Yes. The core monorepo is licensed under the Business Source License 1.1 (BSL 1.1). The source is public, you can read it, fork it, self-host it, and modify it. BSL 1.1 converts to an open-source license (Apache 2.0) after the change date, and the only usage restriction is commercial resale as a competing hosted service.

Question 5

Can I self-host the whole thing?

Accepted Answer

Yes. The public monorepo runs one organization per instance and has no hidden dependency on the SaaS control plane. Clone it, set your environment variables, and run the platform and workspace runtime together. Railway and Render templates are linked in the monorepo README.

Question 6

What is Hierarchical Memory (HMA) and why does it matter?

Accepted Answer

HMA gives every workspace three memory scopes — LOCAL (private to one workspace), TEAM (shared within a parent-child hierarchy), and GLOBAL (org-wide, write-only from the root). Memory sharing follows the org chart instead of leaking across the whole system, so sibling teams stay isolated by default and escalation paths mirror the real hierarchy.

Question 7

What is Hermes and how do skills work?

Accepted Answer

Awareness stores what happened across sessions. Hermes reads that memory back, identifies repeatable patterns, and promotes the ones that work into reusable skills. Skills hot-reload into the live workspace runtime, so the next task benefits from prior success without restarting. This is the self-improving flywheel in the architecture section.

Question 8

What stack is Molecule AI built on?

Accepted Answer

The control plane is Go 1.25 + Gin + Postgres. Workspace runtimes are Python 3.11. The Canvas (the visual org chart + 10-tab ops panel) is Next.js 15 + React Flow + Zustand + Tailwind v4. Secure sandboxing uses four tiers from T1 sandbox through T4 full-host, per workspace.

Question 9

Can I programmatically create and manage orgs via API?

Accepted Answer

Yes. Partner API Keys (GA April 30, 2026) let any CI/CD pipeline, marketplace integration, or automation tool create and manage Molecule AI orgs via a typed REST API — no browser session required. Keys are scoped to the org they create, rate-limited, revocable, and audited. Ephemeral test orgs: POST to create, run your tests, DELETE to clean up. See docs.molecule.ai/architecture/partner-api-keys.

Question 10

What does it mean that Molecule AI has a governance built in?

Accepted Answer

Two Phase 34 features ship governance at the platform level: Tool Trace records every tool call your agents make — name, inputs, output previews, run_id-paired for parallel traces — so you can verify what actually ran. Platform Instructions lets org admins write system-level rules (e.g. 'redact PII before writing to external tools') that are injected into every agent's system prompt at startup, before the first turn. Governance as a first-class platform concern, not a post-hoc filter.

Question 11

How do I get started?

Accepted Answer

Read the quickstart at doc.moleculesai.app, then clone git.moleculesai.app/molecule-ai/molecule-core. The README walks through provisioning the first workspace, picking a runtime adapter, and wiring up your first team. Expect to have a running workspace in under ten minutes.

Question 12

What does it cost?

Accepted Answer

The open-core monorepo is free under BSL 1.1 — you can self-host it forever without paying anyone. The hosted SaaS is live at app.moleculesai.app with signup, orgs, and multi-tenant provisioning on top of the same open core. API available at api.moleculesai.app. Documentation at doc.moleculesai.app.

Question 13

How does Molecule AI handle agent observability in production?

Accepted Answer

Tool Trace embeds a full execution record — every tool call, input, and output preview — in every A2A response. No separate observability stack to integrate, no sampling. For enterprise governance, Platform Instructions lets org admins prepend rules to every agent's system prompt before the first turn — enforced at the platform level, not applied as a post-hoc filter.

Question 14

Can partners or CI/CD pipelines programmatically manage Molecule AI organizations?

Accepted Answer

Yes. Partner API Keys (mol_pk_*) enable programmatic org provisioning and lifecycle management via API — no browser sessions, no manual handoffs. Partners can POST a new org, manage it via API, and DELETE it when done; billing stops immediately on teardown. Keys are org-scoped by design: a compromised key cannot escape its org boundary.

AI agent organizations.

Agent count is exploding

A2A & MCP became standards

Governance is now a buying gate

Runtimes are commoditizing

Fintech · Series A

What I saw

What we agreed

What the org learned

Hermes — the skill distillation layer

Model providers

Runtime adapters

Tools · Eval · Observability

Role > Task

Recursive teams

Org = Topology

Distributed A2A

4-layer security

Tiered HITL

Cross-workspace traces

3-layer Marketplace

moleculesai.app

Control Plane

Plugin

Agent

Org Bundle

Build

List

Earn

Start running an organization.

The operating system for AI agent organizations.

Four shifts converging on multi-agent.