Fly Labs
Ask FlyBot

Open Envelope – an open schema for defining AI agent teams

Built an open JSON Schema for defining AI agent teams.Multi-agent systems are becoming a real deployment pattern — not single assistants, but teams with roles, handoffs, and human checkpoints. But there's no shared way to define one that travels across frameworks. Every implementation is scattered, locked to whichever tool you picked first. Built the schema to fix that.The schema lives at schema.openenvelope.org and is registered in SchemaStore, so if you drop a .envelope.json file in VS Code you get autocomplete and validation without installing anything. It's also on npm as @openenvelope/schema if you want to validate programmatically.The spec covers: agent definitions (role, prompt, model, access policy), supervisor/sub-agent hierarchy, human-in-the-loop gates, pipelines, schedules, and secrets/variables that get injected at deploy time. Access policies let you declare exactly which hosts each agent can call — the runtime enforces this at the network level, not in the prompt.The goal is a portable definition format — define a team once, any compatible runtime can execute it. Similar to how Dockerfiles describe a container without being tied to a specific host. There's a managed runtime at openenvelope.org but the schema is Apache 2.0 and anyone can implement it.Happy to answer questions on any part of the spec — especially interested in feedback from people who've built multi-agent systems and have opinions on what's missing.

Hacker NewsToolAISource
0Sign in to voteCopy link

FL score

72

out of 100

Verdict

VALIDATE

medium confidence

Competition

No competitor data yet

Trend

No signal yet

A portable JSON schema for multi-agent AI systems that lets teams be defined once and run anywhere, similar to Docker for agent orchestration.

The pain

Teams building multi-agent systems today lock themselves into a single framework's definition format. Moving a team from LangChain to Anthropic's framework or a custom runtime requires rewriting the entire configuration. There's no lingua franca for agent teams, so each tool reinvents the wheel.

The gap

The schema exists and is technically sound, but adoption requires either: (1) runtimes implementing it, or (2) developers choosing to use it voluntarily over their framework's native format. Right now it's a standard with no critical mass of implementations. The managed runtime at openenvelope.org is the only execution layer, which creates a chicken-egg problem.

Build angle

Don't lead with the schema as a product. Lead with the runtime. Build openenvelope.org as a managed service that executes .envelope.json files reliably, with better observability/debugging than competitors. Make the schema the implementation detail. Once teams are running production workloads on your runtime, the schema becomes the moat, they can't leave without rewriting. Then license the spec to other runtimes (LangChain, Anthropic, etc.) as a compatibility layer.

Strengths

  • Solves a real coordination problem in multi-agent systems, today's frameworks have no standard way to define hierarchies, handoffs, or access policies
  • Low friction distribution via SchemaStore and npm means developers can adopt the schema without installing anything
  • Access policies enforced at network level (not prompt level) is a genuine security improvement over current practice
  • Apache 2.0 licensing removes vendor lock-in concerns and makes adoption easier for enterprises
  • Timing is good, multi-agent deployments are moving from research to production, and teams are starting to feel the pain of framework lock-in

Risks

  • Schema adoption without runtime adoption is a dead end. If no one implements it, it's just documentation. The creator needs to prove runtimes will actually use it.
  • The managed runtime at openenvelope.org is unproven at scale. If it's slow, unreliable, or expensive compared to alternatives, teams won't use it regardless of the schema's elegance.
  • Existing frameworks (LangChain, Anthropic, Crew AI) have no incentive to support a competing standard. They'll only adopt if forced by customer demand or if it becomes industry consensus, which requires critical mass first.
  • The schema may be missing features that matter in production (error recovery, cost controls, audit trails, model fallbacks). Feedback from actual multi-agent teams is essential before betting on this.
  • Market timing risk: if the multi-agent space consolidates around one or two dominant frameworks before this gains traction, the schema becomes irrelevant.
  • No clear monetization path beyond the managed runtime. If the runtime margins are thin or the market is price-sensitive, the business doesn't work.

Questions about this idea?

FlyBot reads the scoring and gives you a second opinion on “Open Envelope – an open schema for defining AI agent teams”.

Open FlyBot