Skip to content

maturity-model.md

Latest commit

 

History

History
112 lines (77 loc) · 2.97 KB

maturity-model.md

File metadata and controls

112 lines (77 loc) · 2.97 KB
Raw

Maturity Model

Use this model to adopt the skillpack gradually. You do not need every skill on day one, and a small team should not have to carry release-governance weight for tiny local edits.

Optional at every level:

  1. install quizme-mode when you want persistent --quizme exhaustive clarification with optional --mc, --one-at-a-time, --confirm, and --record behavior

Level 1: Baseline Assistance

Goal: make simple work clearer, cheaper, and easier to follow.

Required:

  1. token-reduction
  2. order-of-operations
  3. process-budget-controller

Evidence:

  1. startup declarations use a small skill set
  2. small tasks avoid governance artifacts
  3. final answers stay concise

This level is enough when the main problem is scattered process, not release risk.

Level 2: Developer Workflow

Goal: make normal coding work safer.

Required:

  1. Level 1 skills
  2. scripted-command-execution
  3. diagnose-before-fix
  4. regression-prevention
  5. effective-testing-methods
  6. doc-maintenance

Evidence:

  1. bugs are reproduced before fixes when practical
  2. behavior changes get targeted validation
  3. docs update when workflows change

This is the everyday level for most active repositories.

Level 3: Governed Validation

Goal: make risky or release-affecting changes auditable.

Required:

  1. Level 2 skills
  2. skill-governance
  3. governance-enforcement
  4. semantic-policy-audit
  5. interdependent-change-planning
  6. user-instructions-tracker

Evidence:

  1. governance artifacts exist for governed changes
  2. strict artifact validation passes
  3. CI checks enforce governed-change rules
  4. user directives have evidence rows

Use this level when a future reader needs to understand why a change was safe to push.

Level 4: Lifecycle Management

Goal: keep the skillpack maintainable as it grows.

Required:

  1. Level 3 skills
  2. skill-usage-review
  3. deprecation-management
  4. file-maintenance
  5. artifact-budget-enforcement

Evidence:

  1. overused and underused skills are reviewed
  2. deprecated paths have replacements
  3. cached artifacts stay bounded
  4. stale docs are pruned or refreshed

This level protects the pack from becoming a pile of overlapping process rules.

Level 5: Evidence-Driven Improvement

Goal: improve from real task outcomes, not speculation.

Required:

  1. field notes
  2. task logs
  3. scenario refreshes
  4. usage-review findings
  5. pruning decisions

Evidence:

  1. repeated friction creates concrete improvements
  2. rarely used skills are merged, deprecated, or justified
  3. examples reflect real tasks
  4. validation profiles match actual risk

This is where the pack becomes more useful over time instead of merely larger.

Advancement Rule

Move up a level only when the previous level is being used consistently.

Regression Rule

If users report process friction, drop one level for routine tasks and keep higher levels only for governed work.

The healthiest version of the pack is not the strictest version. It is the version that matches the task.