Just-in-time grants
Access is granted on request and expires on a TTL, with no standing shells.
Just-in-time, recorded terminal access to your clusters
Reach a pod or cluster from the browser through a just-in-time grant that expires on its own. Policy as code decides who can open a session and where, every session is recorded as a tamper-evident replay, and idle or expired sessions are torn down automatically.
The problem
When an engineer needs to reach a pod in production, the path is usually a shared credential, a long-lived config file, or a VPN hop with no record of what happened. Standing access that never expires makes every session a potential exposure, and when something goes wrong there is no replay to show who ran what or when.
Access is granted on request and expires on a TTL, with no standing shells.
Policy as code decides who can open a session and against which target.
Each session is recorded as a replay so what happened is reviewable later.
Idle, expired, or revoked sessions are closed automatically across replicas.
You request a session for a target; policy decides and a grant is issued.
You get a terminal in the portal, scoped to the grant and recorded.
The session ends on its TTL or when idle, and the recording is kept.
How it stays governed
Policy as code governs every access request before a session opens, deciding who can reach which cluster or pod and under what conditions. No terminal session starts without passing that check, and the same rules apply across every registered target.
Each session is recorded as a tamper-evident replay that persists after the session ends. You can show exactly what ran, who ran it, and when, rather than relying on logs that may be incomplete or altered.
When a request requires a person to sign off before a session is opened, the grant waits for that approval rather than opening the terminal immediately.
Works with your stack
Governs access to your existing clusters and pods; IntegraCI adds the policy layer and session recording without replacing your runtime.
Who it’s for
When something breaks in production, engineers open a session through the portal instead of reaching for a shared credential. Policy decides who can connect, the session is recorded from start to finish, and the replay stays on hand for the post-incident review.
Compliance reviews require evidence of who accessed which system and what they did. Just-in-time grants that expire on a TTL replace standing credentials, and the session recording gives auditors a reviewable trace without any extra tooling.
New team members need occasional cluster access but should not hold persistent credentials. Policy as code scopes each session to the right targets, and automatic teardown means no cleanup step is forgotten when the work is done.
No. IntegraCI provides a governed access path through the browser. Your existing cluster runtime keeps running. IntegraCI sits in front of the access request, evaluates it against policy, and records the session without replacing what is underneath.
Rules are written as policy as code by whoever administers your platform. Engineers do not set their own access scope. Each request is evaluated against those rules before a grant is issued, and the policy applies consistently across every registered target.
The recording is kept as a tamper-evident replay that persists independently of the session. You can review it later to see exactly what ran, when, and under which grant. It is not editable after it is written.
Sessions are torn down automatically when the TTL expires or when the session goes idle, across all replicas. There is no manual cleanup step and no standing shell left open after the grant period ends.
Request a demo, or read the docs to see how it fits the tools you already run.