AI code improvement
Coding assistants work from what you tell them. ROOKON starts from what your users experienced (the error, the session, the affected step), gathers the relevant code and has your AI provider propose a focused change, with tests, for your review.
The problem
Generating code is no longer the bottleneck. Deciding which problem deserves attention, assembling the right context and reviewing the change safely still take most of the time.
Live demo 1 of 3
A simulated deck of proposed fixes, each with its evidence, risk, checks and diff. Approve, request changes or save one for later while new requests keep arriving.
rookon · acme-cloud · commit requests
Simulated data#128rookon/fix-legacy-billing-interval → main
How ROOKON helps
The stack trace, breadcrumbs, affected sessions and only the relevant source files are assembled into each analysis, not your whole repository.
Connect OpenAI, Anthropic Claude or Google Gemini with your own API key and choose which model diagnoses issues and which writes the code. Usage is billed by your provider to your account.
Every proposal states what changed, why, the risk involved and how to verify it, so review takes minutes rather than an investigation.
Every code fix comes with tests. Your CI runs them on a draft pull request, and if they fail, Rooky reads why and fixes it. Code without tests or green checks is never offered for Merge.
Changes go to ROOKON’s own branch, never to main. ROOKON never merges on its own: in Co-Rooky, the default, you merge in GitHub, and in Trusted-Rooky or Insane-Rooky it merges only when you tap Merge.
Choose which repositories ROOKON may read. Each analysis only fetches the files it needs, and every access is written to your audit log.
Live demo 2 of 3
Use a fast model for diagnosis and your strongest coding model for the change itself. Model names are examples: ROOKON lists the models your own key can use.
rookon · workspace settings · AI models
Simulated dataConnected providers
Model names below are examples. In ROOKON the list is fetched from your provider account with your key.
| Task | Model | Est. per run |
|---|---|---|
Diagnosis A fast, inexpensive model is usually enough. | ~$0.022 | |
Code generation Pick your strongest coding model. This is where quality matters most. | ~$0.87 | |
Review revisions Defaults to your code generation model. | Uses Code generation: your-strongest-coding-model | ~$0.36 |
Marketing copy Defaults to your diagnosis model. | Uses Diagnosis: your-fast-model | ~$0.006 |
Support replies Defaults to your diagnosis model. | Uses Diagnosis: your-fast-model | ~$0.004 |
Content writing Defaults to your marketing copy model. | Uses Diagnosis: your-fast-model | ~$0.007 |
The model you pick for code generation writes every proposed change. Model routing is in development with the rest of ROOKON.
ROOKON identifies an issue from errors, performance data or user behavior.
It gathers evidence: traces, breadcrumbs, affected sessions and the code paths involved.
Your configured provider receives the context and writes a fix, together with tests for it.
The fix opens as a draft pull request and your CI runs the tests. If they fail, Rooky reads why and tries again.
You see the diff, the tests and the results in one place, and accept or decline.
Your team merges and your pipeline deploys, on your terms and your schedule.
Live demo 3 of 3
A simulated month on your AI key: set a monthly cap and what runs automatically, then see which issues get analyzed, which wait for you and what happens at the cap.
rookon · workspace settings · AI budget
Simulated dataBilled by your AI provider, on your key. ROOKON stops before the cap.
Your model’s price per 1M tokens
Example prices. Check your provider’s current pricing for your model.
Simulated day 1 of 30
Live$0.00 of $30 this month
on trackprojected month-end $0.000 tokens0 automatic · 2 waiting
Detected issues (newest first)
Simulated month. Most impactful issues are analyzed first; ROOKON reserves the worst case before each call so a single analysis can’t push you over the cap.
Try: Drop the budget to $10 · Switch to Full autopilot
FAQ
No. The furthest ROOKON goes is a pull request on its own branch, and only once you allow pull requests. Merging and deployment happen through your own process.
Every code fix comes with tests for the code it changes, and your own CI runs them before you review. If they fail, Rooky reads the failure and tries again, twice by default. ROOKON never offers Merge for code without tests or green checks. No tests in your repository yet? Rooky proposes setting them up first.
ROOKON is built on a provider abstraction, so you choose the provider and, per task, the model: for example a fast model for diagnosis and your strongest coding model for writing the change. ROOKON lists the models your key can access. The integrations and model routing are in development.
Only the files and snippets relevant to the issue being analyzed are sent, and only to the provider you configured. That traffic falls under your own agreement with that provider.
Request changes, close the pull request or dismiss the proposal. Nothing reaches your main branch without a person on your team merging it.
You pay your AI provider directly for the tokens you use. ROOKON does not resell or mark up AI usage, records what each analysis consumed, and stays inside the monthly AI budget you set, so the first month can’t surprise you.
Sign up and follow your first issue from detection to a proposed fix.