initialization-readiness-harness

Separate repo initialization from implementation. Use when a project needs startup readiness, first passing test evidence, task breakdown, scripts/init-check, or a clean initial checkpoint before feat

SKILL.md

GitHub

Initialization Readiness Harness

Purpose

Make the first agent phase establish a runnable, testable, documented baseline before business implementation begins.

Inspect First

  • manifests, lockfiles, README, startup scripts, test setup, local runtime docs, AGENTS.md, CI, and existing task plans
  • whether the repo is new, newly scaffolded, or unreliable to start from scratch

Procedure

  1. Define initialization outputs.

    • docs/development/startup-readiness.md
    • docs/development/task-breakdown.md
    • scripts/init-check when a repeatable check is useful.
  2. Prove the baseline.

    • Install dependencies using the project-native package manager.
    • Confirm the app or service can start, or document the precise blocker.
    • Confirm at least one test path exists and can pass, or document what must be added first.
  3. Separate phases.

    • Do not mix feature implementation into initialization.
    • Initialization completes when a fresh agent can start, test, inspect progress, and pick the first task.
  4. Add advisory validation.

    • Use scripts/check_initialization_readiness.py from this plugin.
    • Keep scripts/init-check focused on readiness, not full release validation.
  5. Route follow-up work.

    • Use feature-state-machine for the task list when work is multi-feature.
    • Use quality-gates-installer for the broader validation interface.
    • Use local-runtime-harness for runtime reset, seed, and healthcheck depth.

Validation

  • Run ./scripts/init-check when present.
  • Confirm startup docs list install, dev, test, validation, ports, and current state.
  • Confirm task breakdown contains ordered tasks with acceptance criteria.

Completion Criteria

  • The repo has a verified initialization baseline or precise blockers.
  • Later sessions can begin implementation without rediscovering setup, test, or task structure.