This page summarises the user-visible features included in AITM 1.0, the first public release, compared with the internal preview builds that preceded it.

What's New

Pipeline heals itself instead of stopping

The biggest change is philosophy: a task should only end as failed when the configuration is wrong or the AI is missing data. Every other problem is now retried or repaired automatically.

• Merge conflicts self-heal. A task that hits a conflict while merging into dev is retried with rebase and merge strategies, then handed to a fix agent.
• Post-merge tests fix forward. If tests fail after a merge, AITM isolates the cause and fixes it on the task branch rather than reverting the merge immediately.
• Type-check auto-fix. A failing type-check goes to a fix agent for up to three rounds before the task fails.
• Infrastructure errors are classified separately from real test failures: a crashed worktree is rebuilt, and dependents are no longer cascade-skipped.
• Watchdog covers every task type, including agent-mode tasks.
• Stale merge locks and stranded merges are reclaimed automatically on the next scheduler tick and after a restart.
• Post-merge test fixes stay within the task's scope. When tests are fixed after a merge, the fix cannot touch files outside the task's scope — the same scope-gate used elsewhere in the pipeline.

Tasks say what they are for

• Intent field. Every task now carries an Intent — bug, feature, enhancement or other. The New Task dialog asks for it (it is pre-filled to bug when you create a task from a bug report), you can change it when editing the task, and tasks created from Chat or Brainstorm get it from the AI. The intent decides which safeguards apply — for example, only bug tasks must prove the bug (see below).
• Chat and Brainstorm split large requests. When you ask for something big, the assistant proposes several smaller tasks with the right intent instead of one oversized task.
• Design simulation for complex tasks. For large or risky changes the Architect step now walks its own plan through a set of "what if" scenarios (empty data, a restart in the middle, two tasks at once, a failing dependency…) before any code is written, and the later steps check the same scenarios. Small tasks skip this automatically.

Merging got stricter — and safer

• Pre-merge build gate. Right before a task is merged into dev, AITM builds and type-checks the result as it will look after the merge. A red result gets a bounded number of automatic fix attempts; if it stays red, the task is not merged. Errors already present on dev are not blamed on the task. See Merge & Review.
• The queue pauses during AI tool updates. AITM now updates the Claude command-line tool itself — at startup and once a day — instead of letting it update in the middle of your work. While an update is pending, running steps are allowed to finish, new steps wait (you see a notification and a small indicator), and the queue continues on its own afterwards.

Creating and checking projects

• Project proposals. The AI (in Chat, Brainstorm or through a connected assistant) can now propose a new project, but never creates one on its own: you get a confirmation dialog where you can edit the details, and nothing is written to disk until you confirm.
• Project health check. Before a project's first task, AITM checks the basics — it is a Git repository with at least one commit, it is on the expected branch, the required files for its technology exist and are valid — and shows any problem as a warning in the task console, instead of letting the task fail half-way with a confusing error. The project panel links to the latest check.
• Repository check on creation. New projects are validated (Git initialized, first commit, required files) at the moment they are created.
• .NET projects set themselves up. When the New Project Wizard onboards a .NET project, it installs the required .NET SDK and the project's dependencies automatically and tells you in real time when it actually installs something.
• Wizard conveniences. The wizard pre-fills the project folder from your projects root (Settings → Pipeline → Wizard projects root), has a folder picker limited to that root, and automatically adds the AITM Dev Panel script to newly created web projects.

Search and everyday comfort

• Search query syntax. Search boxes for variables, contracts, hints and tasks, and the @keyword / %contract suggestions, now understand three operators: a.b — a followed by b in that order; a&b — both, in any order; a|b — either. Matching ignores upper/lower case.
• Reasoning effort. For providers that support it, you can choose how hard the model should think — in Chat, Brainstorm and for pipeline steps. The control is hidden for models that don't offer it.
• Table font slider in Settings → Font & Style sets the row text size of all tables; header and button sizes follow.
• CORS allowed origins can be edited in Settings → System (exact addresses and wildcard patterns) for opening the dashboard from other machines.
• Warnings and notifications all have copy and close buttons.
• Help menu → AITM Dev Panel opens the documentation of the Dev Panel (see Contract Explorer & App Flow).
• The version shown in the dashboard has one consistent format: version · date (build id).

Bug tasks must prove the bug

• Reproduction gate. When a task is a bug fix, the architect step must write a Reproduction section that demonstrates the symptom before the fix and its absence after.
• Diagnostic pack. Bug tasks that AITM files automatically carry the crash lines, a blame summary, affected files and the linked known-error id.
• Business-scope gate. Work outside the task's scope is blocked with an AI judge verdict.

Full test suite is opt-in

Running the whole test suite after every merge was the main cause of memory pressure. It is now off by default (Pipeline → Full test enabled) and, when enabled, runs only when the merge queue is empty.

Deliberate restart after a self-merge

When AITM merges a change to its own code, it no longer restarts under you. The default (Git & Storage → Server restart after self-merge = manual) waits for the queue to drain and shows a Restart button in the header menu.

Settings you could not reach before

• Day / night CPU profiles (System tab): reserve more threads for the dashboard during the day and let tests use the machine at night.
• Disk Janitor and Disk Watchdog (Watchdog & Resources): retention rules and absolute GB budgets per location.
• Swap management (System): recommended swap size computed from RAM and free disk.
• Dedicated Project Check tab, and a Test connection button for the translation model.
• A new project gets a default set of skills right at creation (documentation, UI patterns), instead of waiting for its first task.

Dashboard and panels

• Dockable panels (Chat, Brainstorm, Contract Explorer, task consoles) snap into real layout slots.
• Monitor bar gained a Queued tab, a CPU sparkline, and an Estimation tab with queue ETA.
• English is the default UI language.
• Task cards: dependency numbers are clickable, and the recovery-agent button asks for confirmation.

Licensing

A false-positive licence block can be unlocked from the UI. Project slots used / max are shown in Settings → License. When creating or importing a project in the Wizard, if the licence has a free slot, the wizard offers to assign it right away. Sign in with Google is available both for starting the free trial and for logging in to your account in the licence panel.

Resources

/tmp sweeper recognises leaked Chrome and Playwright directories and cleans them after two hours by default.