People do not usually go looking for a “project ledger.” They go looking for relief from a project problem: work is scattered, a useful chat has gone cold, or the next AI has to be taught the same thing again.
LabNote is for keeping a project coherent across people, chats and AI tools. It is not a replacement for a one-off answer, a plain folder that already does the job, or an automatic agent-memory service.
| If you are thinking… | LabNote gives the project… |
|---|---|
| “I keep losing the thread between chats.” | A selected trail that the next session can read. |
| “I keep repeating the brief to different AIs.” | A shared entry route, handoffs and retained decisions. |
| “A coding agent did work, but the wider project is elsewhere.” | A place to connect plans, reviews, sources and next actions. |
| “I need to know why we decided this.” | Visible sources, review notes and decisions. |
| “I want this simple and under the project’s control.” | Ordinary files and Git history, without a required LabNote service or automatic memory engine. |
The record is deliberately selective. A project can preserve the source, the handoff, the contribution, the review, the decision and the next action without keeping every chat message.
Use the smallest useful tool.
- For a one-shot answer, use the best available assistant and move on.
- For a simple private note, a plain file may be enough.
- For automatic recall inside a running agent, an agent-memory system may fit better.
- For model-call telemetry, token monitoring or runtime tracing, use an observability tool.
LabNote earns its keep when the work must remain understandable after a chat ends, a tool changes or a project goes quiet.