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 takaoumehara/superforge-skill --skill superforge-verifygit 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-verify)<a href="https://agentmods.dev/skills/takaoumehara/superforge-skill/superforge-verify"><img src="https://agentmods.dev/badge/skills/takaoumehara/superforge-skill/superforge-verify/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/takaoumehara/superforge-skill/superforge-verify"><img src="https://agentmods.dev/badge/skills/takaoumehara/superforge-skill/superforge-verify.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00186 | $0.01731 |
| Opus 5 | $0.00093 | $0.00865 |
| Sonnet 5 | $0.00037 | $0.00346 |
| Haiku 4.5 | $0.00019 | $0.00173 |
Grade A, and why
superforge-verify 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 8d 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 — 148 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Superforge Verify — Pre-Completion Verification Gateway
NEVER claim a task is resolved, a bug is fixed, or a feature is complete without gathering empirical runtime proof of success.
0. Know which grade of evidence you are holding
"Paste the real output" is the right rule and only half a rule, because it does not say what makes a piece of evidence good. Evidence that looks convincing while proving nothing is this skill's own failure mode, one level up.
| Grade | What it is |
|---|---|
| A — reproducible | A command anyone can re-run, with its output pasted and the command line above it |
| B — observed | Something captured once: a screenshot at a stated viewport, a log with timestamps |
| C — derived | A conclusion drawn from an A or a B — and it must name which one |
| D — asserted | Someone says so |
A verification report may not contain a single D.
The quiet failure to watch for is a C written in the confident tone of an A.
"Mobile layout verified" is a conclusion; "screenshot at 375px, attached" is
evidence. What makes each grade valid, and the "it worked" versus "it happened
to work" table (cold start is the check most often skipped and the one that
catches the most) → references/evidence.md.
Whoever did the work cannot grade it
Grading your own evidence is a conflict of interest no instruction repairs. The grader has already read the reasoning, already believes the claim, and reads the output looking for the confirmation it expects — which is precisely how a C gets written in the tone of an A.
Where Claude Code is available, run workflows/superforge-verify-evidence.js:
one agent extracts the claims, one runs each proof line and pastes raw output,
and a separate agent grades that output having never seen the implementation
and being asked for the reason it does not prove the claim. Then a reporter
writes docs/verification.md with the mandatory unverified section. Up to
fourteen agents; say "use a workflow" or run /superforge-verify-evidence.
What ships with it
6 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.
- 8d ago First seen · 148 lines · 186 tokens per session scan A af19ba854c63
superforge-verify is a skill published in the GitHub repository takaoumehara/superforge-skill (1 stars, last pushed 23d ago), licensed MIT. It adds 186 tokens to every session and 1,731 once invoked, about $0.0009 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
skillgrade-setup
Sets up and runs skillgrade evaluation pipelines for Agent Skills. Use when initializing eval configurations, running trials, reviewing results, or integrating with CI. Don't use for writing grader scripts, general test authoring, or non-agentic documentation.
onboard-repo
Index an unfamiliar codebase into the knowledge graph, then produce a first orientation map -- entry points, most-depended-upon modules, hotspots, test topology.
intuitive-tests
Use this skill whenever the user asks about unit test best practices, test organization, flat test suites, redundant tests, test refactors, pytest/JUnit/Jest/xUnit layout, test taxonomy, flaky tests, coverage quality, fixtures, mocks, parametrization, pruning existing UTs, or "which tests are worth keeping." It…
lov-fix-until-no-error
A verification-focused debugging workflow that repeatedly runs specified checks, fixes the highest-priority failure, and stops when the checks pass or a genuine external blockage remains.
browser-automation
Browser automation powers web testing, scraping, and AI agent interactions. The difference between a flaky script and a reliable system comes down to understanding selectors, waiting strategies, and anti-detection patterns. This skill covers Playwright (recommended) and Puppeteer, with patterns for testing, scraping…
ox-attest-goal
Pursue a customer capability to proven, Attest-backed BDD: author it from the customer's journey, drive the acceptance run to green, then hand to ox-attest-create to mint the honest red/green proof. Use when a user asks to add, improve, prove, or review a BDD/customer capability, or mentions customer flow, acceptance…