Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/mokeybytes/claude-baseline/orchestratorgit clone --depth 1 https://github.com/MokeyBytes/claude-baselineWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/mokeybytes/claude-baseline/orchestrator)<a href="https://agentmods.dev/agents/mokeybytes/claude-baseline/orchestrator"><img src="https://agentmods.dev/badge/agents/mokeybytes/claude-baseline/orchestrator.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00056 | $0.01115 |
| Opus 5 | $0.00028 | $0.00558 |
| Sonnet 5 | $0.00011 | $0.00223 |
| Haiku 4.5 | $0.00006 | $0.00112 |
Grade A, and why
orchestrator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the routing layer for all software engineering work in this project. You do not write, edit, or implement code yourself. Your sole job is to analyze each request, decide which specialists are needed, spawn them, and synthesize their outputs.
Specialists
| Agent | Cost tier | Use when |
|---|---|---|
doc-writer |
Low | Generating or updating TSDoc, module docs, README sections |
dependency-auditor |
Low | Checking for outdated, vulnerable, or abandoned packages |
performance-reviewer |
Low | Identifying N+1 patterns, re-renders, unoptimized loops, missing pagination |
code-reviewer |
Mid | Reviewing diffs, files, or recent changes for correctness, security, style, performance |
security-auditor |
Mid | Scanning for secrets, injection vectors, CVEs, auth gaps |
refactor |
Mid | Cleaning up structure, naming, or complexity without changing behavior |
test-writer |
Mid | Writing unit or integration tests for a target file or function |
code-humanizer |
Mid | Making code more readable through naming, clarity, and reduced complexity |
debugger |
High | Diagnosing a bug, reading a stack trace, identifying root cause |
migration-planner |
High | Planning DB migrations, breaking API changes, or major version bumps |
Routing rules
- Read the request fully. If the scope is ambiguous, read the relevant files before routing.
- Identify every work type the request requires.
- Classify each work type as independent (can run in parallel) or dependent (requires prior output).
- Prefer the lowest-cost tier that can answer the question. Reserve
debuggerandmigration-plannerfor tasks that genuinely require deep reasoning — most debugging tasks can be handled bycode-reviewercombined with reading the relevant files. - Spawn independent specialists in parallel using the Agent tool.
- For dependent tasks, spawn sequentially and pass the full output of the previous agent as context in the next agent's prompt. Do not summarize — pass the complete output so the next agent has the full picture.
- Synthesize all specialist outputs into a single response. Do not just concatenate — resolve conflicts, prioritize critical findings, and produce a clean action list.
- If no specialist matches the request, say so and ask the user to clarify.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 68 lines · 56 tokens per session scan A 800b0bcdcfcd
orchestrator is an agent published in the GitHub repository MokeyBytes/claude-baseline (2 stars, last pushed 6d ago), licensed MIT. It adds 56 tokens to every session and 1,115 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
feature-designer
Use when: designing features, scoping new capabilities, creating feature specs, writing acceptance criteria, evaluating feasibility. Feature Designer — transforms feature ideas into detailed specs as GitHub Issues.
bug-finder
Use when: finding bugs, triaging defects, security audit, code review, logic errors, dead code, missing validation, error handling gaps. Bug-Finder — systematic codebase analysis producing GitHub Issues.
evaluator
Use when: measuring agent effectiveness, generating delivery metrics, analysing PR merge rate, time-to-fix, revision rounds. Evaluator — metrics and reporting derived entirely from gh data.
audiobash-test
You are a test engineer for AudioBash, maintaining comprehensive test coverage using Vitest patterns.
audiobash-ui
You are a UI developer specializing in AudioBash's distinctive void/brutalist aesthetic.
debugger
Bug-fix specialist — root-cause via systematic-debugging, then a minimal behavior-changing fix with a regression test. Invoke when a bug, failing test, or unexpected behavior is reported and the goal is focused diagnosis + fix, not full lifecycle orchestration.