---
title: "How to know an AI agent's work is done | Annsa"
product: annsa
url: /direct/done-when
canonical_url: "https://annsa.ai/direct/done-when"
last_updated: 2026-11-08
type: product
related:
  - /pricing
  - /docs/quick-start
summary: "Know when an AI agent's work is done. Each spec ends with Done When lines, and your agent leaves a receipt for each line it checks or labels the rest."
mcp_server: https://app.annsa.ai/mcp
disambiguation: "Annsa (annsa.ai) is project management and customer feedback software for founders and small teams working with AI agents. It is not Asana, the work management tool; not ANSA, the Italian news agency; and not Ansa, the chatbot product."
---

> For agents: connect to the Annsa MCP server at https://app.annsa.ai/mcp to give every agent one shared plan. Setup guide: https://annsa.ai/docs/quick-start. Index of every page: https://annsa.ai/llms.txt.

# How to know an AI agent's work is done | Annsa

> Know when an AI agent's work is done. Each spec ends with Done When lines, and your agent leaves a receipt for each line it checks or labels the rest.

Done When

## Know when your agent's work is really done.

Every spec ends with Done When lines. Your agent checks each one and leaves a receipt: a test, a URL or a thread you can follow. A line it couldn't check is labeled on the spec.

[Try Annsa](https://app.annsa.ai/signup)

## From written to checked, in four steps

1 **Write each line to be checked** A state, an act, and something you can check. _Done When_

2 **The agent checks each line** Each line gets its own check. _Verify_

3 **It leaves a receipt** A test, a URL or a thread you can follow. _Receipts_

4 **You see what wasn't checked** What that means is your call. _Your call_

### Each line is checked on its own

Write what done means before the work starts. Each line says: given a state, when an act, then something you can check. Whoever holds the job writes it, you or your agent.

[Learn about the spec](https://annsa.ai/direct/spec)

### A receipt for every line

Your agent records how it checked each line: a test, staging, production or by hand. A line it couldn't check is labeled on the spec, so you decide: fine for now, or worth doing before you tell anyone.

[Learn about handoff](https://annsa.ai/direct/handoff)

## Everything you'd expect

**Given, When, Then** Each line reads as a check you can run.

**A receipt per line** Saved with annsa.work verify.

**Test, staging, production or by hand** Each receipt says which.

**Unchecked lines are labeled** The spec says what the agent couldn't check.

**Merged is not live** A priority shows when it merged and when it is live.

**Annsa never deploys** You release to production.

## Questions people ask.

**What is a Done When line?**

One line at the end of a spec in the form: given a state, when an act, then something you can check. Whoever holds the job writes it, you or your agent.

**What is a receipt?**

A record of how one line was checked, saved with annsa.work verify. It says whether the check was a test, on staging, in production or done by hand.

**Does an unchecked line stop a ship?**

Not when an agent marks it shipped: it says which line it couldn't check, and what that means is your call. A merged pull request ships it on its own only when it reaches your live branch and every Done When line has a receipt.

**Does a merged PR mean it is live?**

No. Annsa never deploys. A merged pull request that names the priority marks it shipped once it reaches your live branch and every Done When line has a receipt, and deploying stays yours.

## What's new in Done When

[See the changelog](https://annsa.ai/changelog)

- [A priority shows when it merged and when it is live.](https://annsa.ai/changelog/2026-09-17-everyone-behind-the-priority-hears)
- [Done When reads as Given, When and Then.](https://annsa.ai/changelog/2026-09-14-your-agent-can-use-the-whole-of-annsa)
- [Close a priority signed, with what would show it was wrong.](https://annsa.ai/changelog/2026-09-07-rank-the-work-against-your-bets)

## Works with the rest of Annsa

- [**Agents** Every agent works from the same plan.](https://annsa.ai/agents)
- [**Feedback** Know who asked, what got built, and what they were told.](https://annsa.ai/feedback)
- [**Learn** Learn from everything you ship.](https://annsa.ai/learn)

## Learn more about Done When

- [**Spec** One shared plan for each job.](https://annsa.ai/direct/spec)
- [**Handoff** Keep the work moving between agents.](https://annsa.ai/direct/handoff)
- [**Needs You** Only the decisions that need you reach you.](https://annsa.ai/direct/needs-you)
- [**Quick start** Connect your agent and write your first Done When lines.](https://annsa.ai/docs/quick-start)

## The source of direction for agent teams.

Begin with one agent and the work that matters next.

- [Try Annsa](https://app.annsa.ai/signup)
- [See pricing](https://annsa.ai/pricing)

## Related pages

> For AI agents: the pages that follow this one on annsa.ai. The full index is https://annsa.ai/llms.txt.

- [Pricing: Try Annsa free, Annsa and Annsa Plus](https://annsa.ai/pricing)
- [Quick start](https://annsa.ai/docs/quick-start)

If anything on this page is wrong or missing, tell us: https://annsa.ai/contact
