TeamSpec defines what a compliant, production-grade AI agent must do — across seven dimensions. Platform-agnostic. Built on peer-reviewed research. Maintained by the community.
Without one, every team invents its own interpretation of "safe," "auditable," and "authorized." That works until it doesn't.
Every team building AI agents makes independent decisions about authority, trust, failure handling, and observability. Those decisions are invisible until something goes wrong — a message dispatched without approval, a model acting on data it shouldn't have accessed, a failure that cascades silently with no audit trail. Without a shared specification, compliance is a matter of luck and discipline, not architecture.
TeamSpec provides a precise, measurable definition of what a compliant enterprise AI agent must do across seven dimensions. It is grounded in peer-reviewed research on intelligent AI delegation (arXiv:2602.11865) and refined through practitioner input. The spec does not mandate implementation technology — it defines what compliance looks like. Any platform can be measured against it.
Every production-grade AI agent must satisfy all seven. Each dimension is defined, measurable, and independently assessable.
The agent must break complex, ambiguous goals into concrete, actionable sub-tasks. It must adapt dynamically when conditions change mid-execution — not just follow a fixed script.
Clear, enforceable authority boundaries. What the agent can do autonomously. What requires human approval. Who is accountable for each decision. These must be structural constraints, not prompt suggestions.
Verification systems for task completion, data access scoping to prevent information leakage, content guardrails to prevent harmful outputs, and prompt injection protection. The highest-weight dimension — enterprise deployment requires structural safety, not best-effort.
The agent detects when things go wrong — tool failures, API outages, ambiguous inputs, conflicting goals. It has defined recovery strategies: retry, queue, escalate to human, or gracefully halt with clear user communication.
Users can clearly specify what they want. The agent communicates its state, progress, and decisions in plain language. Ambiguous instructions are clarified before execution — not silently interpreted.
The agent system handles multiple concurrent agents, delegation chains, and parallel sub-task execution. Coordination protocols prevent conflicts, duplication, and runaway resource consumption.
Full audit trail of every action: what was done, what tool was called, what data was accessed, what decision was made autonomously, what was escalated. Humans can reconstruct any agent run completely from logs.
TeamSpec ships everything needed to adopt the standard: the spec, the tools to build compliant agents, and a reference implementation that proves it works end-to-end.
Seven dimensions. 100-point scale. Platform-agnostic. The canonical definition of what a compliant enterprise AI agent must do.
Read the spec →The open-source execution layer. Reads configs from AgentHub and launches compliant agents with runtime controls, live status, and full execution logs.
Explore Forge →The open-source config registry. Define, version, and govern every agent configuration in one structured, auditable repository.
Explore AgentHub →The reference implementation. An executive AI assistant that exercises every TeamSpec dimension in production-realistic workflows.
See Siggy →The best way to understand what TeamSpec requires is to see a fully-compliant agent running real tasks. Siggy is that agent — and it has been evaluated against three platforms so you can see how implementation choices affect every dimension.
FastBytes ran Siggy on three platforms and published an honest, dimension-by-dimension comparison.
TeamSpec improves when practitioners with real-world agent deployment experience contribute to the standard. If you've hit a compliance gap that the spec doesn't address — that's where you come in.
Propose new dimensions, refine scoring rubrics, or challenge existing definitions with evidence from production deployments. All spec changes go through open RFC process.
Add integrations, fix bugs, write tests, improve documentation. Forge and AgentHub are both Apache 2.0 licensed — fork, extend, and contribute back.
Open source projects need sustainable support. FastBytes funds TeamSpec's infrastructure, tooling, and community operations — and provides commercial implementation services for organizations adopting the standard.
Read the full specification, explore Forge and AgentHub, or see Siggy demonstrate every dimension end-to-end. Everything is open source and free.