Mention HQ¶
Mention HQ aggregates your activity across GitHub, Linear, Slack, Notion and local files — git branches, todo files, markdown docs — and groups it into the subjects you actually work on, so you regain context on a topic fast instead of tab-hopping across ten tools.
It runs locally, against your own accounts. It reads by default and only ever writes to a source on a deliberate action you take — never as a side effect of a sync.
The model, in three words¶
Everything reads as Bucket → Task → Item:
- an Item is one thing from one source — a PR, a Slack thread, a todo line, a branch;
- a Task is a subject you handle, and items attach to it;
- a Bucket is a topic column on the board that groups tasks.
Sources emit items; the engine proposes which task each belongs to; you confirm or skip in catch-up. Your decisions are kept — a sync never overrides them.
-
A PR, a Slack thread, a Linear issue and a local branch about the same subject sit together on one task, in buckets you define.
-
Everything you haven't ruled on in one place. The engine proposes where each item belongs; triage rules auto-skip the noise.
-
PRs carry their review status; local branches show their git-spice stack; a PR joins the branch it was pushed from — the whole stack reads as one card.
-
Next actions, item→task matching and bucket suggestions — from your logged-in
claudeCLI or an API key. Everything works without it. -
One colleague — a Slack id, a GitHub login, an email — collapses to a single person you can merge and give an avatar.
-
Your data is a SQLite file on your machine; secrets live in the OS keychain, never the database or a log line.
Explore without touching your data
task back:seed builds a throwaway demo database populated across every source, so you can
poke around — every screenshot in these docs is captured from it — without ever touching your
real hq.db. See Getting started.