| name | using-wcbs |
| description | Use when starting a WCBS-enabled session or before any project action, including questions, exploration, planning, edits, tests, reviews, or release work. Routes work to the smallest applicable WCBS skill before acting. |
| activation | session bootstrap and any task where a WCBS skill may apply. |
| required_inputs | User request, active runtime identity, skill catalog, and relevant project context. |
| required_outputs | Selected skill route or evidence-first fallback, evidence state, and next action. |
| authority_dependencies | 00_start_here/SOURCE_OF_TRUTH.md; 10_governance/APIVR_EXECUTION_LIFECYCLE.md; runtime_adapters/adapter-registry.yaml. |
| evidence_requirements | A loaded skill, an executed verification, or an honest Verified, Likely, Suspected, Unknown, Not Run, or Blocked state for every material claim. |
Using WCBS
WCBS is an on-demand operating system, not a document dump. Select the smallest
skill that changes the decision or the evidence standard before you act.
Hard Gate
Before a project action, do one of these explicitly:
- Load the relevant skill; or
- State why no skill applies and use the evidence-first fallback below.
User instructions win. Never invent a tool that is absent from the active
runtime map. If the catalog or map is unreadable, stop with Blocked.
Process Router
| Request signal | First skill | Next mandatory skill when code changes |
|---|
| New behavior, feature, workflow, or product idea | brainstorming | writing-plans, then test-driven-development |
| Plan, audit remediation, or multi-step change | writing-plans | test-driven-development |
| Bug, regression, flaky test, crash, or unknown failure | systematic-debugging | test-driven-development |
| Implementation request with approved plan | executing-plans | test-driven-development |
| Independent implementation and review | subagent-driven-development | requesting-code-review |
| Review request or received feedback | requesting-code-review / receiving-code-review | verification-before-completion |
| Branch, PR, merge, or cleanup | finishing-a-development-branch | verification-before-completion |
| New or revised skill, prompt, runbook, or agent | writing-skills | 20-pass-protocol |
Add APIVR and the 16 Elite Goals at the tier that matches risk. Installation,
runtime adapters, external tools, security, or release artifacts are
Comprehensive unless evidence supports a narrower tier.
Evidence-First Fallback
Use only when no listed skill applies: