Orchestration Claude Code skills

Orchestrates large-scale codebase changes: decomposes work into 5-30 independent units, spawns one background subagent per unit in isolated git worktrees, each running tests and opening a PR.

Recurrence primitives: /loop reruns a prompt on an interval (or self-paced autonomous maintenance from .claude/loop.md) while the session is open; /schedule creates cloud routines on Anthropic infrastructure.

Runs independent plan tasks through parallel subagents with two-stage review; 2026 releases cut execution time sharply by replacing subagent review loops with inline self-review.

The biggest agent-harness collection: 278 skills, 67 agents, 15+ hook events, security scanning (AgentShield, 102 static rules) and memory persistence across Claude Code, Cursor, Codex, OpenCode, Zed; flagship skills are continuous-learning-v2 (instinct extraction with confidence scoring), security-scan, eval-harness, e2e-testing, content-engine.

Mega-toolkit combining 135 agents, 176+ plugins, 35 curated skills, 42 commands, 20 hooks and 14 MCP configs, plus a gateway to the SkillKit marketplace; featured pro-workflow plugin does self-correcting memory and parallel git worktrees.

End-to-end feature development workflow with specialized agents for codebase exploration, architecture design, and quality review.

Run Claude in a continuous self-referential loop on the same prompt (the Ralph Wiggum technique) until the task is complete.

Executes implementation plans with independent tasks in the current session using fast iteration and two-stage review (spec compliance, then code quality).

Runs concurrent subagent workflows to parallelize independent chunks of work.

Progressive context refinement for subagents: retrieve, narrow, and re-query instead of dumping everything into context.

Architectures for autonomous Claude Code loops, from sequential pipelines and PR loops to DAG-orchestrated multi-agent systems.

Manage multiple local CLI AI agents via tmux sessions with cron-friendly scheduling.

Drive a self-hosted multi-agent runtime to delegate work across multiple coding CLIs.

Delegate coding tasks (bug fixes, docs, tests) asynchronously to the Google Jules AI agent.

Reruns a prompt or slash command on an interval, or self-paced.

Creates scheduled cloud agents (cron routines) on Anthropic infrastructure.

Operate as an agentic engineer using eval-first execution, decomposition, and cost-aware model routing.

Patterns and architectures for autonomous Claude Code loops : from simple sequential pipelines to RFC-driven multi-agent DAG systems.

Audits Claude Code context window consumption across agents, skills, MCP servers, and rules. Identifies bloat, redundant components, and produces prioritized token-savings recommendations.

This skill should be used when the user asks to "optimize context", "reduce token costs", "improve context efficiency", "implement KV-cache optimization", "partition context", or mentions context limits, observation masking, context budgeting, or extending eff

Design task-local harnesses, eval gates, and reusable skill extraction for Claude dynamic workflow mode and other adaptive agent harnesses.

Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.

Shared orchestration engine for the orch-* skill family. Defines the gated Research-Plan-TDD-Review-Commit pipeline, the size classifier, the agent map, and the two human gates that the orch-* operation skills delegate to. Not usually invoked directly.

Use when the user wants a task done much faster through parallel work, concurrent agents, batched tool calls, isolated worktrees, or many independent verification lanes without losing correctness.

Read a plan document, decompose it into steps, design a per-step agent chain from the ECC catalogue, and emit ready-to-paste /orchestrate custom prompts. Generative only : never invokes /orchestrate itself. Use when the user has a multi-step plan and wants to

Run team-based orchestration for agent squads using work items, ownership, agent Kanban, merge gates, and control pane handoffs.

>- Offers the user an informed choice about how much response depth to consume before answering. Use this skill when the user explicitly wants to control response length, depth, or token budget. TRIGGER when: "token budget", "token count", "token usage", "toke