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/ninemindai/agentgem/agentgem-reportnpx skills add ninemindai/agentgem --skill agentgem-reportgit clone --depth 1 https://github.com/ninemindai/agentgemWrote 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/ninemindai/agentgem/agentgem-report)<a href="https://agentmods.dev/skills/ninemindai/agentgem/agentgem-report"><img src="https://agentmods.dev/badge/skills/ninemindai/agentgem/agentgem-report.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 | $0.00067 | $0.00763 |
| Opus 5 | $0.00034 | $0.00381 |
| Sonnet 5 | $0.00013 | $0.00153 |
| Haiku 4.5 | $0.00007 | $0.00076 |
Grade A, and why
agentgem-report 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 5d 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 — 34 lines — stays where its author put it; the contents beside it link to each section on GitHub.
agentgem-report
You render a session-analysis report for AgentGem: ONE self-contained HTML document built from a FACTS JSON block (a rubric evaluation report — factors, findings, hygiene scores, per-session rows).
Facts are law
Render ONLY what the FACTS block states. Never recompute, extrapolate, average, or invent a number, a session id, or a finding. If a field is absent, omit its section entirely — no placeholders, no "N/A" filler. Your value-add is composition and clarity, not analysis.
Bake the FACTS verbatim into the document as <script type="application/json" id="report-data">…</script> (escape < as \u003c), and render the visible report from that block with inline JavaScript reading document.getElementById("report-data"). Numbers must reach the page through that JSON, not through prose you typed — this is the anti-hallucination seam.
Self-contained, themed
- Inline
<style>and inline<script>only. NO external resources of any kind: no CDN, fonts, images, fetch. Data URIs only if truly needed. - Theme both ways with CSS custom properties: define light values on
:root[data-theme="light"]and dark on:root[data-theme="dark"], defaulting via@media (prefers-color-scheme: dark)when no data-theme is set. - Responsive: relative units, max-width container (~880px), wide tables/charts scroll inside their own overflow-x container. The page body never scrolls horizontally.
Composition (editorial readout, not a SaaS dashboard)
- Structure top-down: eyebrow line (what this is · scope · when) → ONE thesis headline stating the report's single most important truth in plain words → short lede → KPI row (3–5 big numbers with small labels) → one hero visualization or ranked list (the worst/most important item) → supporting cards or a per-item breakdown → a closing "what to do" list drawn from the facts' advice fields.
- Typography: a serif display face (Georgia/serif stack) for the headline, a clean sans (system-ui stack) for body, monospace for ids, paths, and all numbers.
- ONE accent color plus a success green and a warn tone; neutral surfaces otherwise. No gradients, no drop-shadow cards, no emoji, no icon fonts.
- Charts: hand-rolled inline SVG sized from the data (bars, sparklines, a simple curve). No chart libraries.
- Severity reads at a glance: findings that fired lead with their count and severity tone; clean checks collapse into one quiet "passed" line.
- "Passed" and "did not apply" are different results and must never be merged. A check reporting applicableSessions: 0 was never exercised — say it did not apply, never that it passed. A check with a denominator states it ("2 in 9 applicable sessions"), because a rate without its denominator is the thing this report exists to avoid.
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.
- 5d ago First seen · 34 lines · 67 tokens per session scan A e7984de9b4bc
agentgem-report is a skill published in the GitHub repository ninemindai/agentgem (39 stars, last pushed 8d ago), licensed MIT. It adds 67 tokens to every session and 763 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 2026-08-30.
Other skills, from other repositories
pptx
从论文、大纲或结构化文本生成 PowerPoint (.pptx) 演示文稿。Use when 用户需要把一篇论文/文章/大纲做成幻灯片、slides、演示文稿、PPT、deck。Don't use when 只需纯文本总结、生成 Word/PDF、或修改已有 pptx 的单个像素级样式。.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
Read, extract, split, merge, rotate, watermark, fill, OCR, or create PDF files with verification of page counts and text extraction.
docx
Create/edit/inspect Word documents using available managed or host capabilities.
pptx
Create/edit/inspect/verify slide decks and PPTX presentations.
xlsx
Create/edit/analyze/verify spreadsheet workbooks including XLSX/CSV/TSV.