DAAF-Contribution-Community/daaf / .claude/skills/pdf-reading
pdf-reading
VerifiedReads research PDFs with page-anchored provenance. Covers DAAF's deterministic pdf_ingest CLI (unaltered poppler text as the quote anchor, a second extractor, page flags, OCR on scanned pages, 2000 px renders), the lit_verify card gate that checks each verbatim quote and number on its cited page, the vision cross-check policy and its dated vision-validated model list, two-column, table and scan handling, and the arXiv HTML lane (section-cited cards). Use whenever an agent must read, quote, or extract numbers, tables or figures from a PDF already on disk: OA papers saved by web_fetch.sh (exit 8), user-supplied papers, codebooks, methodology reports. For research reading it overrides the synced anthropic-skills:pdf, whose pypdf and pip-install recipes DAAF blocks. NOT for finding or downloading papers (scholarly-search, web-retrieval), or creating, merging or filling PDFs.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 36.0 KB |
| references/arxiv-and-html.md | markdown | 14.5 KB |
| references/evidence-card.md | markdown | 94.4 KB |
| references/extraction-strategies.md | markdown | 9.6 KB |
| references/long-papers.md | markdown | 3.0 KB |
| references/pdf-ingest.md | markdown | 28.7 KB |
| references/toolchain.md | markdown | 8.9 KB |
| references/vision-validation.md | markdown | 47.2 KB |
references/long-papers.md
View on GitHub# Long papers and chunked reading
How to read a document too long for one careful pass: technical reports,
dissertations, government reports over ~60 pages. The rules: one paper per
dispatch; sequential chunks over ~60 pages with a running-notes block, the
Chain-of-Agents pattern (https://arxiv.org/abs/2406.02818, accessed
2026-10-02) [doc]; and the overlap-page lesson of the
user's source skill. The source skill's parallel chunk agents are dropped:
only the orchestrator dispatches, and parallel chunks lose the running notes.
## When to chunk
- Up to ~60 pages: one reader reads the whole paper. A 30-page paper is
about 25–35k text tokens plus ~8 renders at ~4k each [inf].
- Over ~60 pages: the orchestrator splits the paper on page boundaries into
sequential dispatches. The ingest is done **once** for the whole PDF (one
paper key, one manifest); chunks are page ranges of it.
- `pdf_ingest` refuses more than 600 pages by default (exit 7). Raise
`--max-pages` deliberately for a longer document, and say so.
- A reader at ELEVATED context or above returns early with its notes
(CLAUDE.md § Subagent Context Monitoring); the orchestrator re-dispatches
the rest.
## The running-notes block
Each chunk's reader returns, and the next chunk's reader receives, a block
like:
```
RUNNING NOTES (paper KEY, through page P of N)
- Structure so far: sections and the pages they span
- Definitions and abbreviations introduced (page)
- Tables and figures seen, with the page each is on
- Claims recorded so far: claim_ids and pages
- Open threads: a table continued on the next page; a forward reference
- Pages covered so far: 1..P (every page has a coverage entry)
```
Notes summarize; they never stand in for a quote. Every quote in the final
card is copied from `pages/` by the reader who records it.
## Assembling one card
The card covers pages 1..N in one `coverage` array, so the chunks contribute
entries to a single card:
- The orchestrator (or the last chunk's reader) assembles one
`evidence_card.md` from the chunk returns, keeping each entry as written.
- Claim IDs are unique across chunks (prefix by chunk, e.g. `K2.C4`).
- `lit_verify card` runs once on the assembled card; the revision loop
(at most two) applies to the assembled card.
## Overlap pages
A sentence or table that crosses a chunk boundary belongs to the chunk where
it starts. Give the next reader the last page of the previous chunk as
read-only context, and state in the running notes which items continue. A
quote that crosses the page break uses `page_end` like any other span. Do not
record the same page twice: `coverage` rejects a duplicate page (exit 8).
## Very long or scanned documents
Scanned reports (often old federal documents) are slow to OCR (one thread,
300 DPI) and read only as `PASS_OCR` at best, so their cards are capped at
PARTIAL. If the OCR mean confidence in `ingest_manifest.json` is low across
many pages, tell the orchestrator before reading: the paper may not support
numeric claims at all.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| canvas-designanthropics/skillsCreate beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| docxanthropics/skillsUse this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', '.dotx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| pdfanthropics/skillsUse this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting/decrypting PDFs, extracting images, and OCR on scanned PDFs to make them searchable. If the user mentions a .pdf file or asks to produce one, use this skill. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| pptxanthropics/skillsUse this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates (.potx), layouts, speaker notes, or comments. Trigger whenever the user mentions "deck," "slides," "presentation," or references a .pptx or .potx filename, regardless of what they plan to do with the content afterward. If a .pptx or .potx file needs to be opened, created, or touched, use this skill. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |