Cancer Detection API

Detecting cancer from bio-images.

CIAO documentation and architecture sync

This repository can automatically generate a project documentation report and an architecture diagram whenever source changes are pushed. The generated files are committed to the repository root as documentation.pdf and architecture.png by default.

Enable automatic sync

1. In GitHub, open Settings → Actions → General and make sure Actions are enabled.
2. Under Workflow permissions, select Read and write permissions, then save. The workflow requests contents: write so it can commit the generated files.
3. Confirm that .github/workflows/update-ciao-docs.yml is present on GitHub (it is included in this setup). The workflow uses CIAO from 176Anjali/CIAO-Code-In-Architecture-Out-main and runs Ollama on the GitHub Actions runner; you do not need to install Ollama or Docker on your computer for GitHub sync.
4. Push a source-code change, or open Actions → Update CIAO documentation → Run workflow to start it manually.
5. Watch the run under Actions → Update CIAO documentation. The first run can take longer while the runner installs Ollama and downloads the model. On success, download documentation.pdf and architecture.png from the repository root.

The report records the analyzed commit and generation time. Generated commits include [ciao-generated] to prevent a repeat run loop. A failed generation does not replace the last successful files.

Optional repository variables

In Settings → Secrets and variables → Actions → Variables, add any of these if you want different formats or model:

Variable	Allowed values	Default
CIAO_DOCUMENT_FORMAT	pdf, docx	pdf
CIAO_IMAGE_FORMAT	png, svg	png
CIAO_OLLAMA_MODEL	Any model available to Ollama	llama3.2:3b
CIAO_OLLAMA_NUM_PREDICT	Positive token count	2048

The selected report and diagram format are committed to the repository root. The workflow checks out the exact source commit that triggered generation.

Installation

Install the Python packages listed in requirements.txt, then follow the application entry point and run instructions for your local environment. GitHub documentation sync is configured separately by the workflow above.

Project sections

• app/, prediction/, preprocessing/, and training/ contain application and model pipeline code.
• configs/ and models/ contain configuration and model assets.
• dataset/ contains project data.
• tests/ contains automated checks.