---
title: "Task brief"
source: "agents/task-brief.md"
---

> Part of [botgumi docs](https://www.botgumi.com/docs/llms.txt). Every page: [llms.txt](https://www.botgumi.com/llms.txt).

# Task brief

<!-- The orchestrator fills this in for every task. Reviewers check tier, runtime and model against model-policy.md and botgumi.config.ts. -->

- **Issue:** #n (labels `track:*`, `tier:*`, `runtime:*`; `gate` for human gates)
- **Blocked by:** #n, or none
- **Track:** backend | ios | android | landing | dashboard | design | docs
- **Tier:** high | mid | low
- **Why this tier:** one line
- **Worker:** runtime + model, for example `claude · claude-sonnet-5-5`
- **Reviewers:** one (mid, low) or both (high), each on a different runtime from the worker
- **Branch:** `<track>/<topic>`

## Goal

What is true when this is done, in one or two sentences.

## Inputs

- ADRs, contract paths, hi-fi screens, tracking-plan events this task touches.

## Out of scope

## Contract changes

None, or the change request and which tracks need a follow-up.

## Done when

- The track's CI passes, output pasted in **How tested**.
- Strings in en and zh-Hans. All four states handled.
- Docs: the module's pages updated if behaviour, a contract or a design changed; `pnpm docs:check` passes.
- No blocking review findings open; `botgumi/review` is success.
- Merge: by the orchestrator if `merge-policy.md` allows it, otherwise by a human.