thinking-engine
Thinking Engine
Purpose
Explore unclear problems, challenge assumptions, surface blind spots, generate options, and turn ambiguity into a decision-ready plan.
Use When
- The problem is vague or underspecified.
- The user asks what they may be missing.
- Options are needed before implementation.
- A plan needs critique before commitment.
- MVP scope needs validation before execution.
Avoid When
- Implementation is already decided.
- The user wants direct execution only.
- A concise answer matters more than exploration.
More from moeller-projects/agent-skillz
ado-gateway
Use when fetching Azure DevOps work items or pull request comments, parsing Azure DevOps pull request URLs, or normalizing read-only Azure DevOps data for downstream skills. Do not use when editing Azure DevOps state, reviewing GitHub pull requests, or authoring OpenSpec files.
4caveman
Trigger when user wants terse, low-token, high-signal technical replies for coding work. Avoid for documentation, onboarding, polished copy, or high-risk instructions requiring explicit detail.
3ops-engine
Use when designing CI/CD, reviewing infrastructure, working with containers or Kubernetes, or analyzing deployment and security risk. Avoid for pure application refactoring or documentation.
3doc-engine
Use when writing or improving README files, AGENTS.md guidance, architecture notes, or workflow documentation. Avoid when only implementation is requested.
3repo-engine
Use when onboarding into a repository, mapping architecture, finding entry points, or extracting conventions and hotspots. Avoid when exact files are provided and only a small patch is needed.
3test-engine
Use when writing tests, improving coverage, designing E2E flows, or reducing flaky behavior in automation. Avoid when the task is pure refactoring with no testing component.
3