-
Notifications
You must be signed in to change notification settings - Fork 1
working agreement
Shree Chaturvedi edited this page Jun 13, 2026
·
1 revision
This agreement defines how the team plans, builds, reviews, verifies, and documents the Agentic AutoML Platform.
| Role | Owner |
|---|---|
| Product owner / proxy | Shree Chaturvedi |
| Scrum master / facilitator | Ayush Yadav |
| Frontend lead | Zarif Fida Chowdhury |
| Backend/ML lead | Ayush Yadav |
| DevOps/data lead | Shree Chaturvedi |
- GitLab issues and boards are the source of truth for sprint work.
- Issues should include acceptance criteria, test ideas, and relevant dependencies.
- Implementation should happen in focused branches and merge requests.
- Product behavior, setup, API, or architecture changes should update README/wiki/docs as appropriate.
A story is ready when:
- the goal and acceptance criteria are clear;
- dependencies are identified;
- the work can fit in a sprint or has been split;
- the expected test/verification path is known;
- any non-trivial approach has enough design context for implementation.
A story is done when:
- acceptance criteria pass;
- the code is reviewed;
- relevant tests are added or updated;
-
npm run build,npm run test, andnpm run lintpass unless an approved blocker is documented; - related docs/wiki/API notes are updated;
- secrets and sensitive data are not committed;
- the change is merged through the agreed GitLab workflow.
- Prefer strong tests over shallow happy-path checks.
- Use project ownership and deployment ownership middleware for data-bearing routes.
- Keep generated LLM code reviewable and gated by user approval.
- Store secrets in local
.envfiles or GitLab CI variables. - Treat uploaded datasets, documents, model artifacts, and prediction logs as sensitive unless explicitly public.
- Use GitLab for project-traceable work and decisions.
- Use team chat or meetings for quick coordination, then reflect durable decisions in issues, merge requests, or the wiki.
- Surface blockers early, especially environment, Docker, database, LLM provider, or testing issues.
This page lives in the GitLab wiki. Update it when the team's workflow changes materially.