Menu
SkillsMP has collected 11 skills from VKirill/claude-lane-stack. Open a skill to review its source and details.
Showing 11 of 11 collected skills.
Solo file-based multi-lane orchestration. Durable switchable Kimi/Qwen/AGY/Grok controller, no daytime LLM review, nightly Codex review/fix, PM auto-merges to main. Use when planning runs, splitting DAGs, dispatching writers, recovering blocked tasks, or…
Living project memory files: .agents/PROGRESS.md, .agents/LESSONS.md, docs/decisions.md, session-log, agent-notes. Use when finishing a task/session, writing handoff, night audit, onboarding a cold project, or user says progress/lessons/ADR/handoff.
Primary project onboarding for Claude Lane Stack. Diátaxis-classified LLM pack (MODULE_MAP, API_SURFACE, TEST_INDEX, TAXONOMY, RUNBOOK, Google DESIGN.md). Schemas + validation in references/. Weekly docs-maintainer. Use when: /project-onboard, онбординг, init…
OpenAI Codex CLI — OpenAI's official Rust-based agentic terminal coding tool (NOT the deprecated 2021 Codex completion model). Use when: openai codex, codex CLI, @openai/codex npm, codex exec, codex exec review, codex app-server, AGENTS.md,…
Keep ARCHITECTURE/README/PROGRESS (and full-scenario GOTCHAS/TESTING/deployment) fresh for Claude Lane Stack projects after daily code changes. Use when: nightly docs, docs-maintain, обновить документацию, актуализировать ARCHITECTURE.
File-based task contracts under .agents/runs/ with owns_paths, verification tiers L0/L1/L2, and solo merge rules. Use when authoring or reviewing task YAML, owns_paths, acceptance, or verification commands.
Cold-start project context for orchestrator or human. Use when user says resume, продолж, where were we, cold start, or starting a new orchestrator session on an existing repo.
File-based TODO/ideas board for humans + agent technical notes. Use when user says todo, туду, идея, backlog, запиши, занеси, потом, incubate; or when capturing follow-ups from a discussion without starting a run yet.
Hybrid MetaMCP gateway — agentmemory + gitnexus direct; all other MCP tools via MetaMCP discover/call
Anthropic's official Claude Code CLI — terminal coding agent with skills, hooks, subagents, MCP, plan mode, slash commands. Use when: claude-code, @anthropic-ai/claude-code, claude CLI, /init, /agents, /hooks, /mcp, /loop, plan mode, .claude/settings.json,…
Source text: Mixed languages
Behavioral guidelines to reduce common LLM coding mistakes. Use when writing, reviewing, or refactoring code. Derived from Andrej Karpathy's LLM coding pitfalls.