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 TheCraigHewitt/cowork-starter-pack --skill doc-summarizegit clone --depth 1 https://github.com/TheCraigHewitt/cowork-starter-packWrote 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/thecraighewitt/cowork-starter-pack/doc-summarize)<a href="https://agentmods.dev/skills/thecraighewitt/cowork-starter-pack/doc-summarize"><img src="https://agentmods.dev/badge/skills/thecraighewitt/cowork-starter-pack/doc-summarize/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/thecraighewitt/cowork-starter-pack/doc-summarize"><img src="https://agentmods.dev/badge/skills/thecraighewitt/cowork-starter-pack/doc-summarize.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 | $0.00072 | $0.01245 |
| Opus 5 | $0.00036 | $0.00622 |
| Sonnet 5 | $0.00014 | $0.00249 |
| Haiku 4.5 | $0.00007 | $0.00125 |
Grade A, and why
doc-summarize 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 — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Doc Summarize
Turn a 40-page PDF, a 90-minute meeting transcript, or a dense article into a one-page brief the user can actually use.
Before starting
- Ask the user: what's the purpose of the summary? "For my knowledge," "for a decision I'm making," "for a meeting tomorrow" — the answer changes what you emphasize.
- Read
context/preferences.md— citation style, format preferences - If the document is long (>30 pages or >20k words), tell the user roughly how long the summary will take
What to produce
A markdown file saved to /output/summary-{document-name}-{YYYY-MM-DD}.md with four sections:
- The 30-second version — 3–5 sentences. The core takeaway. If the user only reads this, they got the gist.
- Key points — 5–10 bullets. The things that matter. Each one stands alone.
- Notable quotes or passages — verbatim excerpts if the document is worth quoting (reports, transcripts, academic papers). Include page/timestamp references if available.
- Action items or implications — if the summary has implications for the user (decisions to make, things to follow up on), list them. Skip if not applicable.
Inputs
- The document (required) — PDF, markdown, transcript, text file, article URL
- The purpose (recommended — ask if not given)
- Any prior summaries or notes on related documents in the project folder
How it works
- Read or fetch the document. If it's a URL, pull the content. If it's a PDF, extract the text.
- For long documents: make a pass to identify structure (sections, arguments, conclusions), then a pass to extract the key points
- For transcripts: identify speakers, decisions made, action items explicitly stated
- Draft the four sections. Dense, specific, no padding.
- If there are things the document says that contradict other things the user has read (and you have that context), flag it
- Save the file, tell the user one sentence on what surprised you or what they should look at first
What not to do
- Do not summarize by compression ratio ("reduce to 10%"). Summarize by importance — some documents have three things that matter, others have thirty.
- Do not paraphrase every paragraph. Skip the filler. If 80% of a report is methodology and 20% is findings, give them the findings.
- Do not editorialize beyond what the document supports. If you disagree with the document, that's a separate deliverable.
- Do not skip quotes if the document earns them. Some documents are worth reading for the exact language.
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 · 83 lines · 72 tokens per session scan A e7ad595eb41d
doc-summarize is a skill published in the GitHub repository TheCraigHewitt/cowork-starter-pack (5 stars, last pushed 4mo ago), licensed MIT. It adds 72 tokens to every session and 1,245 once invoked, about $0.0004 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
pdf-text-replace
Replace text in fillable PDF forms by updating form field values. This skill should be used when users need to update names, addresses, dates, or other text in PDF form fields.
design-export-repair
Fixes broken decks/PDFs exported from Claude's "Design" feature (or similar AI deck generators) — cut-off or clipped text, wrong/substituted fonts, and corrupted .pptx package structure that shows up only once you export to PDF, not while looking at it in the app. Use this skill whenever the user uploads a zip, .pptx…
google-apps-script
Build Google Apps Script automation for Sheets and Workspace. Custom menus, triggers (onEdit / time-driven / form submit), dialogs, sidebars, email batches, PDF export, external API. Use whenever the user wants to automate a Google Sheet, build a Sheets menu / sidebar / dialog, hit a Sheets row from email or a…
financial-document-parser
Extract and analyze data from invoices, receipts, bank statements, and financial documents. Categorize expenses, track recurring charges, and generate expense reports. Use when user provides financial PDFs or images.
thesis-consistency-audit
A consistency audit for quantitative master’s and doctoral theses in management, finance, or strategy. It checks whether numbers, tables, analyses, and claims agree, and can inspect hidden author information in office documents.
bilingual-paper-reader
A reading tool for translating an English research-paper PDF into Chinese paragraph by paragraph, with color labels for claims, innovations, methods, limits, and quotable sentences. It uses a reusable offline HTML reader and stores highlights in the browser.