Extending User Stories to capture autonomous AI behavior, conditional autonomy, and collaborative intelligence.
An Agent Story extends the User Story paradigm to capture autonomous and semi-autonomous AI behavior. Where User Stories focus on human intent ("As a user, I want..."), Agent Stories must capture emergent behavior, conditional autonomy, and collaborative intelligence.
Agent Stories recognize that modern AI agents operate through stages and state transitions, leverage logical reasoning defined by prompts, maintain memory for learning and context, utilize tools via MCP connections, and respond to triggers ranging from messages to scheduled events.
The format follows a principle of progressive disclosure: the core story remains simple and readable, while structured annotations capture complexity only where it exists.
Every Agent Story starts with this human-readable core. Additional annotations are added only where complexity exists.
AGENT STORY: [ID] As [Agent Role], triggered by [Event], I [Action/Goal], so that [Outcome/Value]. Autonomy: [Full | Supervised | Collaborative | Directed]
Agent Stories come in two formats to match your needs:
For simple agents, early-stage design, and quick stakeholder communication.
For complex agents requiring detailed specifications and engineering handoff.
Events that activate agents: messages (A2A), resource changes, schedules (cron), cascades, or manual activation.
Workflow (predictable stages), Adaptive (runtime decisions), or Hybrid (structured flexibility).
Composable competencies with proficiencies, quality bars, and acquisition types (built-in, learned, delegated).
In-the-loop (every decision), On-the-loop (monitoring), or Out-of-loop (fully autonomous).
Working memory (ephemeral), persistent stores (KB, vector, relational), and learning signals (feedback, reinforcement).
Supervisor (coordinates), Worker (executes), or Peer (collaborates) roles with defined communication patterns.
| Level | Human Role | Agent Authority |
|---|---|---|
| Full | None during execution | Complete decision authority |
| Supervised | Monitors, intervenes on exception | Executes within guardrails |
| Collaborative | Active participant in decisions | Proposes, human confirms |
| Directed | Initiates and guides each step | Executes specific instructions |
Agent Stories focus on what the agent does today. For planning how capabilities evolve over time and how trust develops between humans and agents, use the HAP Plan.