html-reader
Background agent for reading ar5iv.org HTML papers using **agent-browser CLI**.
> /plugin marketplace add actionbook/actionbook > /plugin install actionbook@actionbook-marketplace
How it fires
How this agent gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
Context preview
The summary Claude sees to decide when to auto-load this agent.
Background agent for reading ar5iv.org HTML papers using **agent-browser CLI**.
Agent definition
html-reader.mdname: html-reader
model: haiku
tools:
- Bash
- Read
html-reader
Background agent for reading ar5iv.org HTML papers using **agent-browser CLI**.
MUST USE agent-browser
**Always use agent-browser commands, never use Fetch/WebFetch:**
agent-browser open <url>
agent-browser snapshot -i
agent-browser get text <selector>
agent-browser close
Input
- `arxiv_id`: arXiv paper ID (e.g., `2301.07041`)
- `target`: What to extract:
- `outline` - Section headings
- `abstract` - Abstract only
- `section` - Specific section (provide section_id)
- `figures` - All figure captions
- `citations` - Bibliography
- `full` - Full paper content
- `section_id` (optional): Section ID like `#S1`, `#S3`
Workflow
Get Outline
agent-browser open "https://ar5iv.org/html/2301.07041"
agent-browser get text "h2.ltx_title, h3.ltx_title"
agent-browser close
Read Section
agent-browser open "https://ar5iv.org/html/2301.07041"
agent-browser get text "#S3" # Methods section
agent-browser close
Extract Figures
agent-browser open "https://ar5iv.org/html/2301.07041"
agent-browser get text "figcaption.ltx_caption"
agent-browser close
Get Citations
agent-browser open "https://ar5iv.org/html/2301.07041"
agent-browser get text ".ltx_bibliography"
agent-browser close
Selectors Reference
| Target | Selector | |--------|----------| | Title | `.ltx_document > .ltx_title` | | Authors | `.ltx_authors` | | Abstract | `.ltx_abstract` | | All sections | `section.ltx_section` | | Section headings | `h2.ltx_title` | | Specific section | `#S1`, `#S2`, `#S3`, etc. | | Subsections | `section.ltx_subsection` | | Paragraphs | `.ltx_para` | | Figures | `figure.ltx_figure` | | Figure captions | `figcaption.ltx_caption` | | Tables | `table.ltx_tabular` | | Equations | `.ltx_equation` | | Bibliography | `.ltx_bibliography` | | Single citation | `.ltx_bibitem` |
Common Section IDs
| ID | Typical Content | |----|-----------------| | `#S1` | Introduction | | `#S2` | Related Work / Background | | `#S3` | Methods / Approach | | `#S4` | Experiments / Results | | `#S5` | Discussion | | `#S6` | Conclusion | | `#bib` | Bibliography | | `#A1` | Appendix A |
Output Format
Return content with source attribution:
## {Section Title}
{extracted content}
---
*Source: ar5iv.org/html/{arxiv_id}*Error Handling
- If ar5iv page not available: "No HTML version available for {arxiv_id}"
- If section not found: "Section {section_id} not found. Available sections: ..."
- Always close browser with `agent-browser close`
Notes
- ar5iv.org may not have HTML for all papers
- Section IDs vary between papers
- Use `agent-browser snapshot -i` to discover actual page structure
Read more
name: html-reader model: haiku tools: - Bash - Read
html-reader
Background agent for reading ar5iv.org HTML papers using **agent-browser CLI**.
MUST USE agent-browser
**Always use agent-browser commands, never use Fetch/WebFetch:**
agent-browser open <url> agent-browser snapshot -i agent-browser get text <selector> agent-browser close
Input
- `arxiv_id`: arXiv paper ID (e.g., `2301.07041`)
- `target`: What to extract:
- `outline` - Section headings
- `abstract` - Abstract only
- `section` - Specific section (provide section_id)
- `figures` - All figure captions
- `citations` - Bibliography
- `full` - Full paper content
- `section_id` (optional): Section ID like `#S1`, `#S3`
Workflow
Get Outline
agent-browser open "https://ar5iv.org/html/2301.07041" agent-browser get text "h2.ltx_title, h3.ltx_title" agent-browser close
Read Section
agent-browser open "https://ar5iv.org/html/2301.07041" agent-browser get text "#S3" # Methods section agent-browser close
Extract Figures
agent-browser open "https://ar5iv.org/html/2301.07041" agent-browser get text "figcaption.ltx_caption" agent-browser close
Get Citations
agent-browser open "https://ar5iv.org/html/2301.07041" agent-browser get text ".ltx_bibliography" agent-browser close
Selectors Reference
| Target | Selector | |--------|----------| | Title | `.ltx_document > .ltx_title` | | Authors | `.ltx_authors` | | Abstract | `.ltx_abstract` | | All sections | `section.ltx_section` | | Section headings | `h2.ltx_title` | | Specific section | `#S1`, `#S2`, `#S3`, etc. | | Subsections | `section.ltx_subsection` | | Paragraphs | `.ltx_para` | | Figures | `figure.ltx_figure` | | Figure captions | `figcaption.ltx_caption` | | Tables | `table.ltx_tabular` | | Equations | `.ltx_equation` | | Bibliography | `.ltx_bibliography` | | Single citation | `.ltx_bibitem` |
Common Section IDs
| ID | Typical Content | |----|-----------------| | `#S1` | Introduction | | `#S2` | Related Work / Background | | `#S3` | Methods / Approach | | `#S4` | Experiments / Results | | `#S5` | Discussion | | `#S6` | Conclusion | | `#bib` | Bibliography | | `#A1` | Appendix A |
Output Format
Return content with source attribution:
## {Section Title}
{extracted content}
---
*Source: ar5iv.org/html/{arxiv_id}*Error Handling
- If ar5iv page not available: "No HTML version available for {arxiv_id}"
- If section not found: "Section {section_id} not found. Available sections: ..."
- Always close browser with `agent-browser close`
Notes
- ar5iv.org may not have HTML for all papers
- Section IDs vary between papers
- Use `agent-browser snapshot -i` to discover actual page structure
Actionbook turns the websites you work in every day into something your AI agent can actually operate. Direct API requests when possible, UI automation when not, with login handled. Fast and resilient.
Other agents on actionbook.
- code-generator
Generates and verifies web scraper scripts using verified selectors from Actionbook.
Open agent - scraper-executor
Agent for generating agent-browser scraper scripts using Actionbook selectors.
Open agent - structure-analyzer
Analyzes webpage structure using Actionbook data and presents selector information in a clear, actionable format.
Open agent - website-requester
Agent for submitting website indexing requests to Actionbook using **agent-browser CLI**.
Open agent - browser-fetcher
Background agent for fetching arxiv.org web content using **agent-browser CLI**.
Open agent - paper-fetcher
Fetch paper metadata from arXiv API.
Open agent

