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 matematicsolutions/awesome-matematic-skills-pl --skill doc-intel-contract-plgit clone --depth 1 https://github.com/matematicsolutions/awesome-matematic-skills-plWrote 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/matematicsolutions/awesome-matematic-skills-pl/doc-intel-contract-pl)<a href="https://agentmods.dev/skills/matematicsolutions/awesome-matematic-skills-pl/doc-intel-contract-pl"><img src="https://agentmods.dev/badge/skills/matematicsolutions/awesome-matematic-skills-pl/doc-intel-contract-pl/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/matematicsolutions/awesome-matematic-skills-pl/doc-intel-contract-pl"><img src="https://agentmods.dev/badge/skills/matematicsolutions/awesome-matematic-skills-pl/doc-intel-contract-pl.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 | $0.00297 | $0.03741 |
| Opus 5 | $0.00148 | $0.01870 |
| Sonnet 5 | $0.00059 | $0.00748 |
| Haiku 4.5 | $0.00030 | $0.00374 |
Grade A, and why
doc-intel-contract-pl 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 12d 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 — 209 lines — stays where its author put it; the contents beside it link to each section on GitHub.
doc-intel-contract-pl - kontrakt wyjscia Document Intelligence
Po co
Nasza drabinka PDF (pdftotext -> markitdown -> opendataloader -> Chandra -> vision) daje 5 roznych ksztaltow wyjscia. Ten skill ujednolica je do JEDNEGO kontraktu, ktory od razu odpowiada na trzy pytania:
- Ktory fragment ma zobaczyc czlowiek? - confidence-gating (Article III / AI Act art. 14).
- Co zredagowac? - typed blocks + flagi PII (signature/stamp/PESEL/NIP/IBAN).
- Gdzie w dokumencie jest ten cytat? - bbox + block_id -> most do [[citation-grounding-pl]].
Zamkniety jest MODEL Mistral OCR 4, nie idea. Odtwarzamy kontrakt na wlasnym, lokalnym, RODO-safe stacku. [[feedback_doktryna_skladanie_puzzli_nie_wynajdywanie_kola]]
Kontrakt (v1.1.0)
{
"doc_id": "<sha256 wejscia>",
"contract_version": "1.1.0",
"source": {"path": "...", "engine": "opendataloader|pdftotext|chandra|gaius|vlm-html", "engine_variant": "default|google_doc_ai|null", "pages": N},
"blocks": [
{"id": "b0001", "page": 1, "bbox": [x0,y0,x1,y1]|null,
"block_type": "title|paragraph|table|list|equation|signature|stamp|figure|header|footer|unknown",
"text": "...", "confidence": 0.0-1.0|null, "flags": ["partial","pii_suspected","pii:pesel","sensitive_block","signature_suspected",...]}
],
"gating": {"threshold": 0.85, "review_required": ["b0003"], "auto_approved": ["b0001"]},
"redaction_candidates": ["b0004"],
"meta": {"created_at": "ISO-8601"}
}
bboxznormalizowany 0-1 (przenosny miedzy DPI). Silnik bez bbox ->null+ flagapartial.confidence == null(partial) -> zawsze doreview_required(konserwatywnie; nie wiemy = czlowiek patrzy).
Uzycie (CLI)
cd ~/.claude/skills/doc-intel-contract-pl
python scripts/normalize.py --engine opendataloader wyjscie.json --pretty
python scripts/normalize.py --engine pdftotext dokument.txt --threshold 0.9
cat wyjscie.json | python scripts/normalize.py --engine opendataloader -
Exit: 0 = kontrakt schema-valid, 2 = blad wejscia / kontrakt niepoprawny (pasuje pod CI / pre-commit).
What ships with it
43 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.
- .matematic/konstytucja.md 3.2 KB
- .matematic/spec/001-output-contract-mvp/plan.md 3.5 KB
- .matematic/spec/001-output-contract-mvp/spec.md 3.1 KB
- .matematic/spec/001-output-contract-mvp/tasks.md 3.3 KB
- contract/contract.schema.json 2.0 KB
- LICENSE 1.8 KB
- README.md 3.7 KB
- references/prompt_vlm_ocr_pl.md 3.3 KB
- references/render_skanu_pl.md 1.6 KB
- scripts/adapters/__init__.py 67 B runs code
- scripts/adapters/chandra.py 5.5 KB runs code
- scripts/adapters/gaius.py 1.9 KB runs code
- scripts/adapters/opendataloader.py 2.4 KB runs code
- scripts/adapters/pdf_inspector.py 7.2 KB runs code
- scripts/adapters/pdftotext.py 1.4 KB runs code
- scripts/adapters/vlm_html.py 5.1 KB runs code
- scripts/contract.py 4.5 KB runs code
- scripts/degeneracja.py 1.7 KB runs code
- scripts/grounding_bridge.py 4.9 KB runs code
- scripts/mask_for_model.py 5.0 KB runs code
- scripts/normalize.py 6.2 KB runs code
- scripts/pdfi_extract.py 4.8 KB runs code
- scripts/pii_flags.py 2.9 KB runs code
- scripts/routing_gate.py 13 KB runs code
- scripts/signature.py 2.4 KB runs code
- tests/fixtures/chandra.sample.json 776 B
- tests/fixtures/chandra2.sample.json 612 B
- tests/fixtures/gaius_pl.sample.json 240 B
- tests/fixtures/gaius.sample.json 208 B
- tests/fixtures/opendataloader.sample.json 799 B
- tests/fixtures/plain.sample.txt 188 B
- tests/fixtures/vlm_html.sample.html 913 B
- tests/test_adapter_pdf_inspector.py 5.2 KB runs code
- tests/test_adapters_chandra_gaius.py 5.5 KB runs code
- tests/test_contract.py 4.8 KB runs code
- tests/test_encoding_crlf.py 1.8 KB runs code
- tests/test_grounding_bridge.py 3.2 KB runs code
- tests/test_mask_for_model.py 3.4 KB runs code
- tests/test_pii.py 3.1 KB runs code
- tests/test_routing_gate_abuse.py 3.7 KB runs code
- tests/test_routing_gate.py 4.3 KB runs code
- tests/test_signature.py 2.6 KB runs code
- tests/test_vlm_html_degeneracja.py 4.1 KB runs code
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.
- 12d ago First seen · 209 lines · 297 tokens per session scan A dbc4d7191319
doc-intel-contract-pl is a skill published in the GitHub repository matematicsolutions/awesome-matematic-skills-pl (6 stars, last pushed 18d ago), licensed MIT. It adds 297 tokens to every session and 3,741 once invoked, about $0.0015 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
exam-study-guide
A study-guide builder for a completed exam chapter that has not yet passed its required checks. It creates a structured teaching checklist and, in full mode, a self-contained HTML and printable PDF guide with readable formulas, visible images, explanations, examples, and answers.
exam-cheatsheet
A compiler that turns completed exam notes, mistakes, mastered topics, and reference material into a traceable pre-exam cheat sheet. It can also make a print-ready PDF when visual output or printing is requested.
pdf-cascade
Acquire PDFs for bibliographic references via a strict 8-source cascade (Crossref OA → arXiv → OpenAlex → Unpaywall → HAL → CORE → archive.org → WebSearch queue; optionally Sci-Hub + Anna's Archive, plus Anna's via a windowed browser, in opt-in mode). Each acquired PDF is validated against expected author/title/year…
gemini-file-analysis
Analyze files (PDFs, images, audio, video, large source files) via Gemini's multi-modal file analysis. Use when a file is too large for Claude's context or when the file is a non-text format requiring visual/audio understanding.
invoice-processing
Extract, validate, and classify invoices from PDF, Factur-X, and UBL formats. Multi-country support with field mapping and anomaly detection.
foundry-hosted-agent-validation
Step-by-step process for validating a Python Foundry hosted agent sample (under python/samples/04-hosting/foundry-hosted-agents/) end to end — running it locally (native runtime and azd ai agent run) and after deploying it to an Azure AI Foundry project with azd. Use this when asked to validate a hosted agent sample.