Personal fork of superpowers for strigov: multi-model dev orchestrator on the GPT-5.6 family — Sonnet main thread orchestrates (triage, dispatching, polling, git); an Opus subagent at max thinking effort writes the plan and its authorized revisions; Codex Sol max reviews the plan (loop cap=4, fresh task each round…
Maintain docs/ARCHI.md — a persistent, tool-agnostic architecture memory of the repository. Use when the user asks to create, update, or compact an architecture overview / ARCHI.md, or when repeated Explore passes keep re-mapping the same codebase. Covers creation, the update discipline, the token budget (10-20k…
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Grounded second opinion from Codex on any question — architecture calls, debugging hypotheses, trade-off decisions, red-teaming a conclusion. Advisory only, nothing is gated on the answer. Use when a decision is about to harden and an independent model's disagreement would be valuable, or when genuinely stuck on a…
How to invoke Codex CLI from Claude Code on this machine. Use when any workflow needs to dispatch a Codex task — codex-dispatch wrapper + --background + Monitor polling. Contains the exact wrapper, flags, polling loop, and stale-lock workaround. Read this before the first Codex call in a session, including from…
A workflow for coordinating multiple coding agents during implementation. The main agent organizes the work while assigned subagents handle code, tests, and reviews.
Use ONLY for legacy upstream-style plans — plain markdown without YAML frontmatter, flat Task 1..N structure (no phases). For phased plans with YAML frontmatter (Ф1, Ф2, …), use dev-orchestrator instead — it honors phase boundaries, auto-commits per phase, and routes BE/FE phases to the right implementer.
Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup.
Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation.
Use when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallback.
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always.
A Claude Code plugin for managing legal cases, including intake, cataloguing, legal analysis, hearing preparation, and appeals. Its description is in Russian and identifies it as a fork of Vassal.