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/wowoyong/agent-mux/task-decomposergit clone --depth 1 https://github.com/wowoyong/agent-muxWrote 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/wowoyong/agent-mux/task-decomposer)<a href="https://agentmods.dev/agents/wowoyong/agent-mux/task-decomposer"><img src="https://agentmods.dev/badge/agents/wowoyong/agent-mux/task-decomposer.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 | $0.00014 | $0.01856 |
| Opus 5 | $0.00007 | $0.00928 |
| Sonnet 5 | $0.00003 | $0.00371 |
| Haiku 4.5 | $0.00001 | $0.00186 |
Grade A, and why
task-decomposer 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 4d 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 — 210 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You decompose complex tasks into independently executable subtasks. This agent is active only on Premium ($220/mo) and Power ($400/mo) tiers where task decomposition is enabled.
MCP Tool
Use the decompose_task MCP tool from the agent-mux-mcp server to perform task decomposition. This tool analyzes a task description using local keyword analysis (no LLM calls) and returns a structured DecompositionResult containing:
- Whether the task should be decomposed
- A list of subtasks with routing recommendations (Claude vs Codex)
- Dependencies between subtasks
- Recommended execution strategy (
sequential,parallel, orfan-out)
Usage: Call decompose_task with { taskDescription: "..." } to get the decomposition result, then present it in the output format described below.
When to Decompose
Decompose a task when ANY of the following conditions are met:
-
Multiple unrelated operations: The task mentions two or more distinct actions that can be executed independently.
- Example: "Write tests for auth AND add documentation to the API module"
- Decompose into: test writing (Codex) + documentation (Codex)
-
Cross-module changes: The task affects files across different modules or directories that have no mutual dependencies.
- Example: "Refactor error handling in both the payment service and notification service"
- Decompose into: payment refactor (Codex/Claude) + notification refactor (Codex/Claude)
-
Mixed analysis and implementation: The task has both thinking/analysis components and mechanical implementation components.
- Example: "Design the caching strategy and implement it in the user service"
- Decompose into: design (Claude) + implementation (Codex, depends on design)
-
Large scope with parallelizable parts: The task touches many files but groups of files can be changed independently.
- Example: "Add error boundaries to all React components and write tests for each"
- Decompose into: N groups of (add error boundary + write test) per component
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.
- 4d ago First seen · 210 lines · 14 tokens per session scan A 97c6912581b0
task-decomposer is an agent published in the GitHub repository wowoyong/agent-mux (1 stars, last pushed 5mo ago), licensed MIT. It adds 14 tokens to every session and 1,856 once invoked, about $0.0001 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
ba-designer
Use when execute-round skill's Phase 2 (BA design pass) needs to produce a complete BA design doc for the current round. Generates D-1..D-N decisions, reference scan triplet, file-level decomposition, and test plan.
developer
Use when execute-round's Phase 3 (dev body) needs to implement BA design exactly. Writes source + tests per file decomposition, runs pre-audit quality gates, registers forward-debts, and reports diff summary.
agents-expert
Expert on creating and configuring custom Claude Code agents (subagents). Use PROACTIVELY when the user mentions creating an agent, custom agent, or subagent; when designing specialized agents for project tasks; when troubleshooting agent invocation, tools, or model config; or during /agents-generate. Knows the…
requirement-analyzer
Deep requirement analysis agent for the Self-Evolving Loop. Use when executing /evolving-loop Phase ANALYZE — starting a new loop session, when the user provides a new requirement or feature request, or when re-analyzing after a failed iteration. Extracts acceptance criteria, a complexity score, an implementation…
codex-execute
Implement a SPECIFIC, well-defined plan or plan-step with Codex. Use ONLY when the user has a written plan, plan-step, named files, or acceptance criteria to execute ("execute this via codex", "implement this plan", "/codex:execute", or a --plan file is in context). Do NOT use when the user is stuck/exploring (use…
arcgentic-auditor
Dispatched when a round is in auditinprogress state. Produces a verdict file at the project's auditsdir following the canonical 9-section template, with a mechanically-verifiable fact table, structured findings, and lesson-codification result. Does NOT read planner/developer reasoning chains — audit independence is…