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/takaoumehara/superforge-skill/superforgenpx skills add takaoumehara/superforge-skill --skill superforgegit clone --depth 1 https://github.com/takaoumehara/superforge-skillWrote 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/takaoumehara/superforge-skill/superforge)<a href="https://agentmods.dev/skills/takaoumehara/superforge-skill/superforge"><img src="https://agentmods.dev/badge/skills/takaoumehara/superforge-skill/superforge.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 | $0.00223 | $0.04526 |
| Opus 5 | $0.00112 | $0.02263 |
| Sonnet 5 | $0.00045 | $0.00905 |
| Haiku 4.5 | $0.00022 | $0.00453 |
Grade A, and why
superforge 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 6d 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 — 346 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Superforge — Concierge & Orchestrator
Routes the work, assigns the model tier, keeps the artifacts. The specialist skills own their craft; this skill owns the order of operations and the fact that nothing gets lost between them.
0. First run — ask once, then never again
This suite is written in English. The person using it may not be.
On the first invocation in a project, before anything else: check whether
docs/superforge.md exists. If it does, read the language settings and follow
them silently — never ask again.
If it does not exist, and the request is substantial enough to be worth a setting (skip this entirely for a one-off question or a two-line fix), ask one question with your inference already filled in:
初回だけ確認させてください。(このスキル一式は英語で書かれています)
**話す言葉**: 日本語 ← あなたの書き方から推測しました
**docs/ に残すファイルの言葉**: 日本語
**非同期エージェント(Devin など)**: 使っていない ← 推測
そのままでよければ「はい」。変えるなら番号で:
[1] 両方とも英語で
[2] 会話は日本語、ファイルは英語(海外のチームと共有する場合)
[3] 別の言語 — 言語名を書いてください
[4] 非同期エージェントを使っている — 名前を書いてください
以後は聞きません。`docs/superforge.md` に保存し、変えたくなったら
「言語を変えて」「Devin を使う」などと言ってください。
Five rules that keep this from being an annoyance:
- Infer first, then confirm. The language the user just wrote in is the answer nine times out of ten. Asking an open question whose answer is already on screen reads as not paying attention.
- Ask it in the inferred language, not in English. A question in English is itself a wrong answer to "what language do you want".
- Offer the split. Conversation and artifacts are genuinely different choices — a Japanese maker with an international repository often wants Japanese replies and English files, and no one thinks to ask for that.
- Never ask twice, and never block. If the user ignores the question and states their task, take the inference, record it, and get on with the work.
- Put every once-only question in this one. The tools line belongs here and
nowhere else — asking "do you use Devin?" three weeks later, at the moment
someone wants a thing built, is an interruption rather than a service (§6,
superforge-handoff/references/external-agents.md§0).
What ships with it
12 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.
- assets/templates/superforge.md 273 B
- README.es.md 3.9 KB
- README.ja.md 4.3 KB
- README.ko.md 4.1 KB
- README.md 3.7 KB
- README.zh-CN.md 3.6 KB
- references/artifacts.md 5.8 KB
- references/help.md 10 KB
- references/intake.md 3.7 KB
- references/model-prompting.md 10 KB
- references/run-log.md 6.7 KB
- references/wiring.md 14 KB
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.
- 6d ago First seen · 346 lines · 223 tokens per session scan A d5b3061cf286
superforge is a skill published in the GitHub repository takaoumehara/superforge-skill (1 stars, last pushed 21d ago), licensed MIT. It adds 223 tokens to every session and 4,526 once invoked, about $0.0011 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 skills, from other repositories
spec-kitty-implement-review
Orchestrate the implement-review loop for Spec Kitty work packages using any configured agent. Covers agent dispatch, state transitions, rejection cycles, arbiter escalation, and dependency-aware sequencing across all 13 supported coding agents. Triggers: "implement and review WPs", "run the implement-review loop"…
spec-kitty-mission-system
Understand how Spec Kitty missions work: the 4 built-in mission types, how they define workflows via step contracts and action indices, how missions and work packages relate, how templates are resolved through the 6-tier chain, and how doctrine artifacts (procedures, tactics, directives) compose mission behavior.…
spec-kitty-program-orchestrate
Orchestrate a multi-repo, multi-mission Spec Kitty program end-to-end: run specify → plan → tasks → implement → review → merge → mission-review → post-merge fixes across several repositories in a defined dependency order, using background sub-agents for parallel work and a pulse-heartbeat safety net for long…
spec-kitty-runtime-review
Review runtime-owned outputs using the Spec Kitty review workflow surface, then direct approval or rejection with structured feedback. Triggers: "review this work package", "check runtime output", "approve this step", "review WP", "is this WP ready to approve", "check this implementation". Does NOT handle: setup-only…
spk-run-verdict-capture
Set a Spec Kitty work-package review verdict through the deterministic event-log seam, so every agent harness records approve/reject the same way.
spk-run-implement-review
Orchestrate Spec Kitty work-package implementation and review loops until all packages are done, approved, or correctly rejected.