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 skills/hashangit/openfusion/skillnpx skills add hashangit/openfusion --skill skillgit clone --depth 1 https://github.com/hashangit/openfusionWhat 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 | $0.00116 | $0.01481 |
| Opus 5 | $0.00058 | $0.00740 |
| Sonnet 5 | $0.00023 | $0.00296 |
| Haiku 4.5 | $0.00012 | $0.00148 |
Grade A, and why
openfusion 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 2d 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 — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OpenFusion
OpenFusion is your panel of expert reviewers, not another worker. You are the executor: you read the code, run the tools, gather the evidence, and do the implementation. When you hit a genuinely hard judgment — one where a single model's answer isn't trustworthy enough and being wrong is expensive — you bring a prepared dossier to OpenFusion and get one consolidated answer back. Then you act on it.
A fusion call fans your prompt out to several models in parallel and a judge reconciles their answers into one (consensus, contradictions resolved, blind spots surfaced). It's ~2–3× slower and costlier than a normal call, and it returns one answer — so it rewards calling it once, well-prepared, over calling it repeatedly.
The mental model
- You do the legwork. Read files, run searches, reproduce the bug, write the code, gather sources. OpenFusion does none of this — it has no tools and sees only what you pass it.
- Bring a dossier, not a question. A good fusion call hands the panel everything a senior reviewer would need: the relevant code, the error/trace, what you've tried, the constraints.
"how do I fix my bug?"is a bad call; the bug + the failing code + the trace + your hypotheses is a good one. - One call per hard problem. Don't loop on fusion to incrementally work something. Prepare, ask once, read the answer, proceed. Don't call it again to validate its own answer.
Pre-flight gate (run this before every fusion call)
Call fusion only when all three are true:
- I've already gathered the concrete material (code/diff/error/reproduction/sources) — or the task is genuinely pure reasoning that needs no external input.
- A single capable model probably isn't enough — the decision is subtle, contested, high-stakes, or benefits from independent perspectives.
- The stakes justify the wait — being wrong is costly (production bug, irreversible action, architecture you'll build on).
If any is false, don't call fusion: do the work yourself, or answer directly.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2d ago First seen · 76 lines · 116 tokens per session scan A 1dec0a441142
openfusion is a skill published in the GitHub repository hashangit/openfusion (34 stars, last pushed 24d ago), licensed MIT. It adds 116 tokens to every session and 1,481 once invoked, about $0.0006 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 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…