ivan-amiqa/claude-code-agent-orchestration
Five Claude Code sub-agents: one boss, cheaper specialists, and a script to measure what it saves you.
stack: TypeScript
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
GUARDRAILS
3/5
Blocks destructive commands · Pre-tool screening hook · No YOLO mode · details
Quality 30/100 Guardrails v2
Copy this rig
# review before running: this installs third-party code
$ npx degit ivan-amiqa/claude-code-agent-orchestration/.claude ./rig-claude-code-agent-orchestration # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ curl -fsSL --create-dirs -o .claude/agents/implementer.md https://raw.githubusercontent.com/ivan-amiqa/claude-code-agent-orchestration/main/.claude/agents/implementer.md $ curl -fsSL --create-dirs -o .claude/agents/log-digger.md https://raw.githubusercontent.com/ivan-amiqa/claude-code-agent-orchestration/main/.claude/agents/log-digger.md $ curl -fsSL --create-dirs -o .claude/agents/pr-reviewer.md https://raw.githubusercontent.com/ivan-amiqa/claude-code-agent-orchestration/main/.claude/agents/pr-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/scout.md https://raw.githubusercontent.com/ivan-amiqa/claude-code-agent-orchestration/main/.claude/agents/scout.md $ curl -fsSL --create-dirs -o .claude/agents/verifier.md https://raw.githubusercontent.com/ivan-amiqa/claude-code-agent-orchestration/main/.claude/agents/verifier.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"permissions": {
"deny": [
"mcp__ccd_session__start_session",
"mcp__ccd_session__hand_off_to_session"
]
},
"hooks": {
"PreToolUse": [
{
"matcher": "Bash",
"hooks": [
{
"type": "command",
"command": "bash \"$CLAUDE_PROJECT_DIR\"/.claude/hooks/worktree-guard.sh"
}
]
},
{
"matcher": "Bash|Read|Grep|Glob|Agent|Edit|Write|NotebookEdit",
"hooks": [
{
"type": "command",
"command": "bash \"$CLAUDE_PROJECT_DIR\"/.claude/hooks/delegation-guard.sh"
}
]
}
]
}
} Subagents (5)
| implementer model: claude-sonnet-5-5 | Executes an already-decided plan — writes the code, runs typecheck and tests, iterates until green. Use it for mechanical and well-specified work: applying a reviewed plan, boilerplate, test writing, |
| log-digger model: claude-haiku-5-5 | Triage high-volume output — cloud logs, container logs, test suite runs, build failures, long SQL result sets — and return only the signal. Use whenever the raw output would be more than ~200 lines, s |
| pr-reviewer model: claude-opus-5-5 | Reviews a feature-branch diff before it merges. Use when a branch or PR is ready for review. Reads the diff against the correct base, checks the project's invariants, and reports findings ranked by se |
| scout model: claude-haiku-5-5 | Read-only exploration and file discovery. Use it to locate where something lives, trace a flow end to end, enumerate call sites, or answer "which file/module handles X" — before any planning or editin |
| verifier model: claude-opus-5-5 | Independent fresh-context check of a claim or a finished change — "is X actually fixed?", "does this really work for a normal user?", "is this doc still true?". Runs with no memory of how the conclusi |
Hooks (3)
| event | matcher | runs |
|---|---|---|
| PreToolUse | Bash | bash "$CLAUDE_PROJECT_DIR"/.claude/hooks/worktree-guard.sh |
| PreToolUse | Bash|Read|Grep|Glob|Agent|Edit|Write|NotebookEdit | bash "$CLAUDE_PROJECT_DIR"/.claude/hooks/delegation-guard.sh |
| PostToolUse | * | bash "$CLAUDE_PROJECT_DIR"/.claude/hooks/context-size-guard.sh |
Permissions
deny (2)
mcp__ccd_session__start_session
mcp__ccd_session__hand_off_to_session
ask (0)
—
allow (0)
—
Similar rigs
leavemagic-cyber/earned-confidence
A behavior contract for AI coding agents — every rule added after a real failure. Ships with a verifier agent that cannot edit what it judges.
Pragmatist 3.6k tok · · Q5
doublegate/Master-Claude
Curated, project-agnostic synthesis of AI-agent knowledge across a multi-project workspace, plus POSIX-sh tooling to install and keep it in sync inside any project. One source of truth for Claude Code, Codex, and Gemini (AGENTS.md + symlink
Pragmatist 985 tok · · Q0
datenwissenschaften/retro-speedlab
Cookiecutter template for Retro Speedlab game packages: a skeleton that scheduled Claude Code lab runs grow until the TAS search beats the game.
YOLO Cowboy 353 tok · · Q6
j0yen/dotfiles
The claude-self cluster: machine-personal CLAUDE_SELF.md, its SessionStart hook, and the claude-self CLI.
YOLO Cowboy 6.6k tok · · Q10