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 skills add humanerd-drew/opencode-drewgent --skill fusion-agentgit clone --depth 1 https://github.com/humanerd-drew/opencode-drewgentWrote 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/skills/humanerd-drew/opencode-drewgent/fusion-agent)<a href="https://agentmods.dev/skills/humanerd-drew/opencode-drewgent/fusion-agent"><img src="https://agentmods.dev/badge/skills/humanerd-drew/opencode-drewgent/fusion-agent.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 5.1 | $0.00068 | $0.01529 |
| Opus 5 | $0.00034 | $0.00764 |
| Sonnet 5 | $0.00014 | $0.00306 |
| Haiku 4.5 | $0.00007 | $0.00153 |
Grade A, and why
fusion-agent 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 — 147 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Fusion Agent — Multi-Model Deliberation on OpenCode Go
Concept
Fusion은 여러 관점에서 동시에 분석한 뒤 구조화된 취합을 거쳐 최종 답변을 만드는 패턴. OpenRouter는 이걸 인프라 레벨(panel + judge API)로 제공하지만, OpenCode Go에서는 같은 패턴을 task()로 구현 가능 — 구독료 안에 포함되어 추가 비용 없음.
flowchart LR
request[Research Question] --> orchestrator[You / Orchestrator]
orchestrator -- panel: parallel delegate_task --> P1[Explorer: perspective A]
orchestrator --> P2[Explorer: perspective B]
orchestrator --> P3[Explorer: perspective C]
P1 --> judge[Judge: structure analysis]
P2 --> judge
P3 --> judge
judge -- consensus / contradictions / blind spots --> orchestrator
orchestrator --> answer[Final Answer]
When to Use
| Use Fusion | Don't Bother |
|---|---|
| Research questions: "Compare X and Y" | Short tactical prompts: "What's the weather?" |
| High-stakes: code review, architecture decision | Simple Q&A with a known answer |
| Expert critique: "What are the strongest arguments for/against?" | One-line implementations |
| Ambiguous: multiple valid answers, trade-offs | Fact lookup (single source of truth) |
| Cost of being wrong > cost of extra completions | Time-sensitive: Fusion adds latency |
Since OpenCode Go is subscription ($0/call marginal), cost is never the reason to skip Fusion — only latency matters.
How to Run Fusion (3 Phases)
Phase 1: Panel — Parallel Deliberation
Spawn N subagents (2-5), each with a distinct perspective. Use explorer profile (or implementer for code tasks) — flash-tier is fine since the judge catches blind spots.
task(subagent_type="explorer", description="Security analysis", prompt="Analyze this question from a security perspective: <question>")
task(subagent_type="explorer", description="Performance analysis", prompt="Analyze this question from a performance perspective: <question>")
task(subagent_type="explorer", description="Maintainability analysis", prompt="Analyze this question from a maintainability perspective: <question>")
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 · 147 lines · 0 tokens per session scan A 092e5dc38c0e
fusion-agent is a skill published in the GitHub repository humanerd-drew/opencode-drewgent (2 stars, last pushed 1mo ago), licensed MIT. It adds 68 tokens to every session and 1,529 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 2026-09-03.
Other skills, from other repositories
itchio-new-game-hunt
Hunt for fresh browser-playable games on itch.io /newest that are worth building SEO arbitrage sites for. Crawls new releases, scores them by signals, produces ranked shortlist.
code-refactoring-assistant
Suggest and apply code refactorings to improve readability, maintainability, and code quality. Use this skill when improving existing code structure, eliminating code smells, applying design patterns, simplifying complex logic, extracting duplicated code, renaming for clarity, or preparing code for new features.…
function-class-generator
Generate complete, production-ready functions and classes from formal specifications, design descriptions, type signatures, or natural language requirements. Use this skill when implementing APIs from specifications, creating data structures from schemas, building classes from UML diagrams, generating code from…
scenario-generator
Generate comprehensive test scenarios, user stories, and acceptance criteria from requirements and specifications. Use this skill when planning testing efforts, writing user stories, creating test cases, documenting acceptance criteria, exploring edge cases, designing test coverage, or translating requirements into…
specification-generator
Generate formal specifications including preconditions, postconditions, invariants, and contracts from code or requirements. Use this skill when documenting APIs, creating formal verification annotations, defining function contracts, specifying class invariants, writing design-by-contract code, or preparing code for…
technical-debt-analyzer
Detect and analyze areas with high maintenance cost, poor design, or accumulated technical debt. Use this skill when reviewing codebases for quality issues, planning refactoring efforts, conducting code audits, assessing project health, identifying maintenance hotspots, or prioritizing technical improvements. Analyzes…