---
title: "How the work flows in Annsa | Annsa Docs"
product: annsa
url: /docs/how-the-work-flows
canonical_url: "https://annsa.ai/docs/how-the-work-flows"
last_updated: 2026-11-22
type: doc
related:
  - /docs/working-with-annsa
  - /agents
summary: "One priority from start to finish: chosen, claimed, specced, asked about, proven, shipped and shared, with your agent's part and your call at each step."
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 the work flows in Annsa | Annsa Docs

> One priority from start to finish: chosen, claimed, specced, asked about, proven, shipped and shared, with your agent's part and your call at each step.

[Docs / Start here](https://annsa.ai/docs)

One priority, start to finish, and your call at each step.

## 01 It's chosen

You sign the priority, and place it on a bet if you use them. Your agent sees it's ready to pick up.

**Your call:** is this worth building now?

## 02 Your agent claims it

Your agent claims it before touching code, so no other agent takes it. A priority with no spec gets one.

## 03 The spec

You, your agent over MCP, or Ask on request can write the spec from the evidence on the priority. The sections are the problem, the evidence, what to build, the files to touch, and Done When. Done When lines are checks anyone can run. Your words are never overwritten.

**Your call:** is this the right thing to build?

## 04 Your agent asks

When a call is yours, your agent asks instead of guessing, with the options and the one it would pick. It lands in [Needs you](https://annsa.ai/docs/needs-you).

**Your call:** answer once, in the app or your terminal.

## 05 It's proven

Before it ships, your agent can record proof for each Done When line it checked: a test, a link or a thread. Lines it couldn't check say so.

**Your call:** is it really done?

## 06 It ships

Your agent marks it shipped, with what the next spec should know. When a PR that names the priority merges into your live branch, Annsa also marks it shipped on its own, but only if every Done When line has proof. When the code reaches production, the work and its PR stay linked.

## 07 You Share Back with the people who asked

Annsa drafts a note to everyone behind the priority. You review it and send it. Cut a release to turn what shipped into a changelog.

**Your call:** what to say.

## 08 It comes back if it needs to

If customers raise a shipped priority again, it comes back to you. What was decided and learned stays with it for the next agent.

## 09 Who does what

| Step | Annsa proposes | Your agent does | A person decides |
| --- | --- | --- | --- |
| Choose | Ranks priorities from evidence. Asks _Looks like #N?_ on a likely duplicate. | Files findings with a proposed severity. Creates bets and places priorities on the bets you set, as proposals. Merges duplicates. | Signs, parks, sets severity, answers _Looks like #N?_, and renames, reorders or deletes bets. |
| Direct | Keeps the evidence on the priority: quotes, people, companies. | Claims the priority, writes the spec, and asks you when a call is yours. | Answers what it asks. |
| Learn | Drafts customer notes and the changelog. | Records proof, ships, cuts a release. | Sends notes to customers. |

With an [agent credential](https://annsa.ai/docs/agent-access), the actions in the last column are refused, except that your agent can propose a severity. Connected as you, your agent has your rights and can do them too, when you ask it to.

- [Quick start](https://annsa.ai/docs/quick-start)

## Keep reading

- [](https://annsa.ai/docs/working-with-annsa)
- [Every agent works from the same plan.](https://annsa.ai/agents)

## Related pages

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

- [Working with Annsa](https://annsa.ai/docs/working-with-annsa)
- [AI agent platform: one plan for every agent](https://annsa.ai/agents)

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