Tasks
Work that survives the session.
Turn scattered chat requests into structured, trackable, resumable work that humans and agents share.
What it is
wikiTaTa turns AI-assisted work into structured, trackable execution. Tasks can be assigned, organized, resumed, audited, and connected directly to project knowledge and operational workflows. Instead of vague chat requests scattered across sessions, work becomes persistent and coordinated — making it easier for teams and AI agents to collaborate without losing progress or duplicating effort.
How it works
Tasks are first-class records
Each task has a title, status, assignee, priority, and links to the cards that describe the work. It is not a chat message or a sticky note — it persists, can be queried, and survives session boundaries.
Cards and tasks stay linked
A task is attached to the project knowledge it belongs to. An agent working on a bug reads the related architecture card before acting, and marks the task done when the work is verified — keeping knowledge and execution in the same place.
Shared ledger for humans and agents
Both people and AI agents read and update the same task list. There is no separate human board and agent queue — work is coordinated in one place, so effort does not get duplicated and nothing falls through the gap between the two.
Audit trail per task
Status changes, assignee updates, and completion events are recorded with timestamps. You can reconstruct exactly what happened on a piece of work, who touched it, and in what order — without relying on anyone's memory of the session.
What it changes for you
- Work agreed on in one session is waiting, assigned, and ready when the next session starts — no reconstruction from chat history.
- AI agents and teammates work from the same list, so you stop discovering duplicated or conflicting effort after the fact.
- Vague chat requests become specific, trackable items with a clear status — either done or explicitly not yet done.
- Reviewing what got done last week means querying the task ledger, not scrolling through transcripts.