Your laptop restarted. Your session didn’t die
Every session is recorded as it runs. Reopen it, read the recap, continue where the agent stopped — on any machine, even the one at home.
Ship faster with code you understand and trust
kcap captures your coding-agent sessions in real time so you can revisit decisions, review the work, and bring that context into your next task
Stay where you already work
Your agents. Your tools. Your team’s memory
Shared knowledge for everything your agents build
Record every coding-agent session, see why the code changed, and start the next session with what your team already learned
Interactive preview · Sample data · Click to explore · Hover for detailsInteractive demo · Sample data
Speed demands
staying unstuck
Out of Claude credits at 4pm? Continue in Codex at 4:02
kcap is vendor-neutral. The session’s context follows you across thirteen harnesses, so switching tools doesn’t mean starting over.
“What was I working on?” — answered, with receipts
Ask what you did yesterday and get the actual sessions back. What To Do Next ranks what’s in flight, with the receipts, and Work on this starts an agent on it.
Your agent made that same mistake last Tuesday. Make it the last time
Evals score sessions against your rubric, and what they learn is injected into future runs. Mistakes become guardrails.
Where do my tokens actually go?
Auditable charts of usage and spend — per repo, per agent, per person. Know before the invoice does.
Weekly spend by repository
northwind/paymentsnorthwind/checkoutnorthwind/infra
Capture, inspect, share and reuse
the work you do with coding agents
Keep what it took
to write the code
Your corrections and unfinished work survive the session
kcap keeps the recorded prompts, tool calls and results alongside the work. Return to the failed attempt, the changed approach or the check still left to run.
You shouldn’t have to know what you’ll need in a month to keep it today
Review the code
and the decisions behind it
A second agent. The same recorded context
Give a reviewer the decisions, changes and checks behind the work. Follow its findings back to the evidence and see what still needs another look.
One agent builds. Another reviews
Give the next session
a better starting point
Earlier work changes how your agent's next session tackles the problem
Recall the failed approach and the team’s existing solution before writing the next fix. Keep useful knowledge within reach of the people and agents working on the project.
Build on what already worked
A growing body of work
Distinct threads that stay connected
More sessions. More agents. More people
Keep the intent and experience that connect the work
Illustrative timeline from installing kcap through Day 1, Day 10 and Day 30+, across Claude Code, Codex, Cursor, Copilot and Gemini. Four independent project threads—Payments, Orders, Customer portal and Deployment—move between agents while retaining distinct colors and shapes. Eight dotted arrows show learning reused across projects, with three labeled examples: Retry safely from Payments to Orders, Validate inputs from Orders to Customer portal, and Reuse checks from Customer portal to Deployment. Pulses follow these arrows during the animation and highlight where learning is received. Projects stay separate. All underlying session and reuse marks remain visible.
- Recorded sessions
- Reused context
- Shared learning
- Payments
- Orders
- Customer portal
- Deployment
Start with the work you have
Bring in existing sessions. Keep the decisions, attempts and changes behind the work.
Keep the intent in view
Follow meaningful threads across people and agents, with the evidence behind each decision still connected
Carry experience forward
Finished work leaves more to build on: approaches to reuse, mistakes to learn from and context for the next project
The work keeps growing
So does the experience you can draw on
Before you ask
What is kcap?
Shared memory for coding agents. kcap records every Claude Code, Codex, Cursor, Gemini CLI, Copilot CLI, Kiro, OpenCode, Pi, Antigravity, Muse Code, Grok Build, and Mistral Vibe session, then makes that record searchable, shareable, and usable by the next agent — so you can continue lost work, switch tools without starting over, recall prior decisions, and review PRs with the reasoning.
kcap keeps a searchable record of your coding-agent sessions. Revisit decisions, review what happened, and bring that context into the next task.
Which coding agents does it work with?
Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot CLI, Kiro, OpenCode, Pi, Antigravity, Muse Code, Grok Build, and Mistral Vibe. Recording is opt-in per agent. Claude Code, Codex, Cursor, Gemini CLI, Copilot CLI, Kiro, OpenCode, Pi, Antigravity, Muse Code, and Grok Build support hosted launches from the dashboard when installed and authenticated on a connected daemon; controls depend on the vendor and platform.
My agent already has session history
Per tool, per laptop. Claude Code can resume Claude Code — on that machine. kcap is one record across thirteen harnesses and your whole team, and it’s served back to every agent, not just the one that wrote it.
How long until this is actually useful?
About five minutes. kcap import backfills your existing sessions, so search and recap work before you’ve recorded anything new.
You’re storing my transcripts?
Only if you turn recording on, and never for projects you exclude. Local keeps your session record on your machine. Team stores it in a shared workspace and includes managed evaluations.
We already have observability tools
Those measure your app in production. kcap records how your software gets built — the agent sessions behind the commits.