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/smileynet/teach-me/visual-qanpx skills add smileynet/teach-me --skill visual-qagit clone --depth 1 https://github.com/smileynet/teach-meWrote 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/smileynet/teach-me/visual-qa)<a href="https://agentmods.dev/skills/smileynet/teach-me/visual-qa"><img src="https://agentmods.dev/badge/skills/smileynet/teach-me/visual-qa.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 | $0.00054 | $0.01756 |
| Opus 5 | $0.00027 | $0.00878 |
| Sonnet 5 | $0.00011 | $0.00351 |
| Haiku 4.5 | $0.00005 | $0.00176 |
Grade A, and why
visual-qa 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 yesterday.
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 — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Visual QA
Verify that UI components render and behave correctly by exercising them and analyzing the evidence.
General Check
Run the automated tool to exercise all components across all pages:
mise run visual-qa
This produces .scratch/visual-qa/manifest.json + screenshots per page. If it exits 0, all behavioral checks pass (tooltips appear, trays open, quizzes give feedback, SVGs render). If it exits 1, something is broken — read the manifest for which checks failed.
The tool is a behavioral check. It answers "does this work?" not "does this look right?"
Navigation Journey (per-domain)
mise run visual-qa exercises components on a page. To verify the cross-page USER JOURNEY
(does clicking through actually navigate?), run the navigation suite:
mise run test:nav
It discovers all library domains from the aggregate index #page-data island (no hardcoded
slugs), self-serves the library/ root headless, and for EACH domain walks
aggregate → domain map → a lesson → its quiz → breadcrumb back-nav, plus the index resume
cue. Navigation is asserted by act-then-verify (click → URL changes → landed <h1>), not by
link-presence. Per-domain pass/fail + screenshots land in test-results/ (navigation-report.md
screenshots/nav-*). Exit 0 = every domain's journey navigates correctly.
NOT in core mise run verify (slower browser journey) — run it after nav/breadcrumb/map/quiz
changes, or when adding a domain. The two-view Tree|Map toggle + tree keyboard model are
covered separately by mise run verify's interactive gate (index_two_view_toggle,
index_tree_keyboard).
Feature-Specific Visual Review
After building or modifying a specific feature, run the tool with --focus to scope screenshots, then analyze those screenshots against the feature's design intent.
python tools/visual-qa.py --serve --focus glossary
Then load the screenshots and analyze. The analysis prompt should be tailored to what the feature is supposed to look and feel like.
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.
- yesterday Changed · +21 lines 25fda432bb96
- 5d ago First seen · 130 lines · 54 tokens per session scan A 086a1e12ac3e
visual-qa is a skill published in the GitHub repository smileynet/teach-me (3 stars, last pushed today), licensed MIT. It adds 54 tokens to every session and 1,756 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 2026-08-31.
Other skills, from other repositories
anki
Anki spaced repetition learning system reference. Covers the science of SRS and SM-2 algorithm, card design principles, optimal deck settings, FSRS scheduler, study workflow, essential add-ons, custom templates, filtered decks, and AnkiConnect API.
reading-metaskill
当用户想养成阅读习惯、问「读什么书/怎么读」「如何学习新领域/怎么入门某学科」时调用。 核心理念: 阅读是终极元技能; 读你所爱直到爱上阅读, 没有读完义务; 读原著与经典优先; 以教促学; 每天1-2小时即可进入极少数人行列。 不适用于: 具体某本书的书评、考试备考资料选择。 Triggers: 阅读/读书/怎么学习/入门/原著/书单/reading/how to learn.
coach
Learning telemetry, strategy, and schedule — retention stats, calibration, grader audit, n-of-1 experiments, HTML dashboard. Use for "how am I doing", weekly check-ins, strategy questions, auditing the grader, or adjusting how Engram teaches.
textbook-distillation
Turn a textbook or long-form source into a self-paced learning track: intake the material, build a chapter map, draft a lesson plan, then generate self-contained HTML lecture notes in a style the human specifies (layout, palette, emphasis), each lesson carrying worked examples, exercises, and checkpoint questions.…
classify-interview-questions
将批量面经或零散面试题逐题去重并分发：Agent/LLM/AI工程题写入 zero2Agent 的 learn-agent-interview，传统后端八股写入相邻 zero2Leetcode 的夏季八股。大批量输入使用 gpt-5.6-luna API 逐篇并发抽题和语义召回，再审查、去重和写答案；不新建面经实录文章。.
new-article
在 zero2Agent 项目中创建新的学习文章。当用户说"写一篇新文章"、"创建文章"、"新建文章"、"在某模块下添加一篇关于X的文章"、"帮我起草一篇讲XX的内容"、"整理面经"时触发。适用于所有模块下新建内容，包括面试维度拆解文章和面经实录。即使用户没有明确说"文章"，只要涉及给 zero2Agent 项目增加教学内容，也应当触发此技能。.