AI CodeOps

AI CodeOps changelog: what's new

What changed in AI CodeOps, newest first. Follow it as RSS, Atom or JSON Feed.

  • Signed evidence packs for every change

    Download a signed JSON record of how a change was made: task, models, approvals, findings, tests and delivery. Anyone can verify it without an account.

  • An AI Help assistant in every workspace

    A Help button on every workspace page opens an assistant that answers from the Help Center guide and your workspace's facts, and hands over to support.

  • Support tickets, with every reply approved by a person

    Open a support ticket from the Help panel. Our support desk uses AI to draft replies, and a person reads and approves each one before it is sent to you.

  • A review the AI could not complete now holds the change

    If the AI findings review returns nothing readable or runs over budget, the change is recorded as unreviewed and the release gate holds it until it is reviewed.

  • Claude, Kimi and GLM models for AI sessions

    AI sessions can now run on Anthropic's Claude, Moonshot's Kimi and Z.ai's GLM models, on the platform's keys or with a provider key your workspace connects.

  • Model tiers, monthly allowances and Auto

    Sessions on the platform's models now run on Economy, Standard or Premium, each with a monthly token allowance. Auto picks the right tier for each step.

  • Prompt caching for Claude sessions

    Claude sessions now use prompt caching, so the conversation history in each turn is read from cache at a fraction of the input price, and usage counts that.

  • A staging release and a browser test suite

    A staging copy of the platform now runs beside production, and a real browser walks sign-up, sign-in and the public pages against it before a build moves on.

  • Sign-in and MFA hardening

    Sign-in now binds each login to your browser with PKCE, MFA is enforced on every server action, and changing a second factor needs a fresh MFA check.

  • Two-person review for AI changes

    Once a workspace has a second person who can approve, the member who launched an AI session can no longer approve its code change alone. A solo team still can.

  • Passkey sign-in and recovery code fixes

    Accounts with only a passkey can verify with it, passkeys without a PIN now work, enrolling a factor issues recovery codes, and a sign-in loop is fixed.

  • Terminal sessions in their own sandbox

    The browser terminal and the step that runs your tests now use their own isolated sandbox, with no access to our systems, private networks or cloud metadata.

  • The release gate judges the latest finished run

    A passing re-run now clears an earlier failure, a run still in progress never counts as a pass, and tests only count for the version of the code they ran on.

  • CodeOps Studio is now AI CodeOps

    CodeOps Studio is now AI CodeOps, at aicodeops.dev. codeopsstudio.dev redirects page for page, and nothing changes for your workspaces, projects or keys.

  • Ollama Cloud as a session provider

    Workspaces can now connect an Ollama Cloud API key and run AI sessions on its hosted open models, including gpt-oss, Kimi K2.7 Code, DeepSeek V4 Pro and GLM.