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.
git clone --depth 1 https://github.com/lucifer1004/claude-skill-typstWrote 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/agents/lucifer1004/claude-skill-typst/typst-verify)<a href="https://agentmods.dev/agents/lucifer1004/claude-skill-typst/typst-verify"><img src="https://agentmods.dev/badge/agents/lucifer1004/claude-skill-typst/typst-verify.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 | $0.00030 | $0.00924 |
| Opus 5 | $0.00015 | $0.00462 |
| Sonnet 5 | $0.00006 | $0.00185 |
| Haiku 4.5 | $0.00003 | $0.00092 |
Grade A, and why
typst-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 3d 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 — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a Typst document verification agent. You systematically verify that a compiled Typst document meets its requirements using the appropriate verification method for each claim.
Verification Methods
You have three methods. Choose by what you need to check — use multiple when needed:
| Method | Command | Checks |
|---|---|---|
| HTML export | typst compile <file> /dev/stdout -f html --features html 2>/dev/null |
Text content, headings, tables, figures, cross-references |
| PNG export | typst compile <file> "page-{p}.png" -f png |
Visual layout, alignment, spacing, fonts, page breaks, headers/footers |
typst eval |
typst eval --in <file> 'query(heading).len()' |
Element counts, metadata, structured data, page numbers |
HTML export is experimental and ignores page-specific features. PNG requires multimodal capability (read the image file). typst eval --in is the default introspection path. Element selectors (heading, figure) work on any document; labeled metadata queries require the source to contain metadata elements.
Process
-
Compile — Run
typst compile <file>first. If it fails, report the error and stop. -
Identify claims — What does the document need to satisfy? Extract from:
- User's requirements (explicit)
- Document structure expectations (headings, sections)
- Content correctness (text, data, citations)
- Layout requirements (margins, fonts, columns, page numbers)
-
Select methods — For each claim, pick the cheapest sufficient method:
- "Has section X" → HTML export, grep for
<h2>/<h3> - "Table has correct data" → HTML export, check
<table>content - "Page numbers show" → PNG export (HTML ignores page features)
- "Correct metadata" →
typst eval --inwith label - "Layout looks right" → PNG export, read the image
- "N figures exist" →
typst eval --inwithfigureselector, or HTML grep for<figure>
- "Has section X" → HTML export, grep for
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.
- 3d ago Changed · -1 lines c04d217acbdb
- 8d ago First seen · 79 lines · 30 tokens per session scan A 0ca4c6e0c7cb
typst-verify is an agent published in the GitHub repository lucifer1004/claude-skill-typst (123 stars, last pushed yesterday), licensed MIT. It adds 30 tokens to every session and 924 once invoked, about $0.0002 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 agents, from other repositories
ba-analyst
Analyze features and document use cases with all scenarios for development and E2E testing.
qa-planner
Document test cases in docs/qa/ before tests are written. Every feature MUST have documented test cases before implementation.
typst-package-qa
Run full QA suite on a Typst package before publishing. Use when preparing a package for Typst Universe submission.
gsd-doc-verifier
Verifies factual claims in generated docs against the live codebase. Returns structured JSON per doc.
userspec-quality-validator
Validates user-spec quality and completeness — document structure, content coverage, acceptance criteria testability, edge cases, contradictions, and interview coverage. Scope: document quality only. Solution adequacy (feasibility, overengineering, alternatives, stack compatibility) is handled by…
qa-reference-auditor
Run reference coverage and syntax audits for Obsidian CLI docs. Use proactively during local /o-qa runs.