What traceknot actually tracks
Every prompt, chat turn, tool call, and subagent — attributed to the work item you claimed, rolled up into a real cost figure.
Inside a session
Every prompt, chat turn, tool call and subagent, broken down by model, tokens and cost. Click into any step for the full picture.
Session
Arguments
node_8fcf54acd5
Sample data, for illustration
Not just what it cost, where it went
Open any session and see which step ran long, where the agent stalled or looped, and what each part actually cost.
Cost per task, not per month
Every session rolls up into a real figure, broken down by model, cache tier and tool calls.
Tied to your issue tracker
Attribute sessions to an issue in your tracker, using the CLI you already have signed in. Read-only, no setup.
Drill into any session
Step-by-step view of where a session actually spent its time and your money.
What the numbers are actually for
A per-session cost figure on its own is just trivia. Attributed to a work item and compared across sessions, it starts answering real questions.
Which task quietly cost 10x the others
Per-task rollups surface the outlier ticket, instead of one blended monthly number that hides it.
Where a session stalled or looped
The step-by-step view shows a tool call repeated or a turn that ran long before you even open the transcript.
How agents or models actually compare
Same kind of task, tracked the same way, whether it ran in Claude Code, Codex, or Copilot — so a comparison means something.
What never got claimed
Sessions that ran without a work item attached stay visible as unclaimed, instead of disappearing into a blended total.
Want the "why" behind a number, not just the number? →Ask your agent, from its own history.