Coders for Humanity
⌘ Shared context. Human contribution.
HANDBOOK / PROJECT LIFECYCLE

From problem to stewardship

Building is a decision supported by evidence, not the automatic next step.

1. Proposal

Identify the problem, people affected, existing solutions, evidence, technological rationale, risks, expertise, and proposed outcome.

2. Research

Validate the need, assumptions, existing interventions, domain constraints, regulatory considerations, and feasibility. Reject or return proposals when evidence does not justify the next step.

3. Specification

Define requirements, users, architecture, data, privacy, accessibility, security, deployment, and success measures. Publish unresolved questions.

4. Build

Break the specification into contribution-sized work units with acceptance criteria, dependencies, and named review responsibility.

5. Verification

Complete automated tests, security review, accessibility review, documentation review, and relevant domain review. Unresolved material risks block a pilot.

6. Pilot

Operate in a limited, consenting real-world setting with rollback and support plans. Collect only necessary data.

7. Measurement

Compare actual outcomes with the proposed baseline and success measures. Publish uncertainty and unintended effects.

8. Deployment

Expand only where evidence supports it and operating responsibility is clear.

9. Stewardship

Maintain security, documentation, governance, infrastructure, contributor succession, and an eventual retirement plan.