llm-plugins-fusion

Generate a Java/Spring backend design plan with mandatory 12 sections and write to PLAN_OUTPUT_PATH.

Run a semi-automated Codex review -> Claude Code fix -> local checks -> Codex verify closure loop for the current branch. Use when Claude Code should orchestrate review-driven fixes with external Bash scripts, structured review artifacts, validation reports, merge readiness, and residual-risk summaries.

Run Codex-only branch review with external scripts and structured review artifacts. Use when Claude Code should collect current branch diff context and ask Codex to produce a high-confidence review report without modifying code.

Run Codex verification against an existing review.md and the current branch state. Use when Claude Code or a human has already applied fixes and only needs a focused verify pass, merge recommendation, and residual blockers.

Lightweight observer-style exploration for quick understanding alignment.

Reviewer-style exploration focused on questions and risk signals, without proposing fixes.

Unified exploration Hub Skill. Route by PERSPECTIVE to observer/reviewer style outputs; analysis only, no design or implementation.

Minimal close-out summary: what changed, why, and known limitations.

Finalize completed work artifacts. Produce commit/PR text in Git repo, else local handoff summary and manual steps.

Fast pragmatic implementation for small tasks; allows minor adjustments but avoids overengineering.

Implement strictly from an approved plan. Requires PLAN_INPUT_PATH and PLAN_APPROVED=true before execution.

Controlled implementation from confirmed steps/plan with minor corrective adjustments only.

Lightweight planning skill for quick execution alignment; non-formal and no code writing.

Critically review a plan for decision clarity, assumptions, and execution risk without rewriting it.

Write a formal plan document to file using general or java-backend profile; design checkpoint only.

Quick lightweight review for obvious, high-signal issues in day-to-day PR checks.

Standard-depth review for correctness, performance, concurrency, failures, tests, and maintainability.

Unified review Hub Skill. Route by LEVEL to lite, standard, or strict review outputs; no code modification.

Exhaustive high-stakes review for production-critical code including boundary/security/data integrity concerns.

Read-only workflow routing skill. Use when a request needs the next nova command, skill, core agent, capability packs, required inputs, and validation path selected before work starts.

Deep exploration skill for complex requirements/incidents; can export analysis artifact identical to chat output.