Skip to content
Skills Library
Documents & office · Analysis

longread

/longread

Use this skill when an agent (main agent or subagent) encounters a file too large to read in a single pass — e.g. cat, read_file, or Read tool hits size limits or truncates output. First assess whether the file is suitable for chunk-based parallel reading, then proceed accordingly. Supports PDF, DOCX, TXT, MD, PPTX files. NOT for structured data (CSV, DTA, XLSX, etc.).

Repository
kimi-skills ↗
Licence
MIT
Last repository change
17.07.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
2
Scripts
1
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 17.07.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What to check before installing
  • 1 scripts in this package. Review them before installing, as Claude may execute them.

Scripts: scripts/split_doc.py

What it does and when to use it

Use this skill when an agent (main agent or subagent) encounters a file too large to read in a single pass — e.g. cat, read_file, or Read tool hits size limits or truncates output. First assess whether the file is suitable for chunk-based parallel reading, then proceed accordingly. Supports PDF, DOCX, TXT, MD, PPTX files. NOT for structured data (CSV, DTA, XLSX, etc.).

Arguments

/longread <file_path>
Original author description

Use this skill when an agent (main agent or subagent) encounters a file too large to read in a single pass — e.g. cat, read_file, or Read tool hits size limits or truncates output. First assess whether the file is suitable for chunk-based parallel reading, then proceed accordingly. Supports PDF, DOCX, TXT, MD, PPTX files. NOT for structured data (CSV, DTA, XLSX, etc.).

How it works

  1. Step 0: Assess Suitability (REQUIRED)
  2. Workflow (only after confirming suitability)
  3. Script Options
Original SKILL.md headings · 76 lines, 404 words
SKILL.md preview
---
name: longread
description: >
  Use this skill when an agent (main agent or subagent) encounters a file too
    large to read in a single pass — e.g. cat, read_file, or Read tool hits size
    limits or truncates output. First assess whether the file is suitable for
    chunk-based parallel reading, then proceed accordingly. Supports PDF, DOCX,
    TXT, MD, PPTX files. NOT for structured data (CSV, DTA, XLSX, etc.).
user-invocable: true
allowed-tools: Bash, Read, Agent, Glob
argument-hint: <file_path>
---

# Longread Skill

Use this skill when a file is too large to read in a single pass (e.g. cat, read_file, or Read tool hits size limits or truncates output).

## Step 0: Assess Suitability (REQUIRED)

Before splitting, determine whether the file is actually suitable for the chunk-and-summarize pattern. **Not all large files benefit from this approach.**

### Files SUITABLE for this skill (non-structured, prose-like content):
- PDF documents (reports, papers, books, manuals)
- DOCX documents (articles, contracts, essays)
- TXT / MD files (long-form text, documentation)
- PPTX files (slide decks with text content)

### Files NOT suitable — use code instead:
- **CSV, TSV, DTA, XLS/XLSX** — structured/tabular data. Use pandas, Stata, or other data tools to query, filter, aggregate. Splitting rows across chunks destroys data integrity.
- **JSON, JSONL** — structured data. Use jq or Python to parse and extract.
- **Log files** — typically need grep/awk/filtering, not summarization.
- **Source code files** — use grep, AST tools, or targeted reads with offset/limit.

### Also consider whether the task itself fits the pattern:
- **Suitable tasks**: summarization, information extraction, question answering over prose, finding specific sections in a long document.
- **Unsuitable tasks**: statistical analysis, counting, aggregation, joins, sorting, exact search — these need code, not parallel reading.

**If the file or task is unsuitable, do NOT proceed with this skill.** Instead, use the appropriate tool (Python/pandas for data, grep for logs, targeted Read with offset for code, etc.) and tell the user why you chose that approach.

---

## Workflow (only after confirming suitability)

### Step 1: Split the Document

```bash
python /app/.agents/skills/longread/scripts/split_doc.py <file_path>
```

The script will output JSON with chunk file paths:
```json
{
  "status": "success",
  "chunk_files": ["/mnt/agents/chunks/doc_part_1.txt", ...],
  "num_chunks": 5
}
```

### Step 2: Create a Reader Subagent

```
create_subagent(
  name="chunk_reader",
  system_prompt="You are a document reader. Read the assigned chunk carefully and extract key information. Summarize the main points concisely."
)
```

### Step 3: Launch Parallel Tasks

For each chunk file, launch a subagent in **parallel** (single message, multiple tool calls):

```
task(agent="chunk_reader", prompt="Read /mnt/agents/chunks/doc_part_1.txt and summarize the key points.")
task(agent="chunk_reader", prompt="Read /mnt/agents/chunks/doc_part_2.txt and summarize the key points.")
...
```

### Step 4: Aggregate Results

After all subagents complete, combine their summaries to answer the user's original question.

## Script Options

The split script supports:
- PDF, DOCX, TXT, MD, PPTX files
- Default chunk size: 32k tokens with 10% overlap
- Output directory: `/mnt/agents/chunks/`
Files in this package (2)
  • SKILL.md 3 KB
  • scripts/split_doc.py 10 KB

Versions

The database has no other skills named longread.

Database snapshot: 13.11.2026 · package version d2a634cd8aa9 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 95Documents
    Read, write, convert, and analyze documents — routes to PDF, DOCX, XLSX, PPTX sub-skills for creation, editing, extraction, and format conversion. USE WHEN document, process file, create document, convert format, extract text, PDF, DOCX, XLSX, PPTX, Word, Excel, spreadsheet,…
  • 87ttokttak
    Create, convert, read and edit Office documents locally with ttokttak-workspace's command line. Build a new presentation (pptx) from a brief through a checked outline, per-page spec and render pipeline, a new spreadsheet (xlsx) from CSV or JSON data with formulas, a new Word…
  • 95heavy-file-ingestion-claude-code
    Use in Claude Code when a user asks to read, analyze, summarize, or extract from a heavyweight file such as PDF, DOCX, PPTX, XLSX, CSV, or TSV. Convert the file into markdown or CSV first with the bundled script, generate a lightweight index, and only spend model tokens on the compressed artifact.
  • 95heavy-file-ingestion-codex
    Use in Codex when a user asks to read, analyze, summarize, or extract from a heavyweight file such as PDF, DOCX, PPTX, XLSX, CSV, or TSV. Convert the file into markdown or CSV first with the bundled script, generate a lightweight index, and only spend model tokens on the compressed artifact.
  • 100markitdown
    Convert documents to Markdown using MarkItDown MCP tools. Use when the user wants to convert PDF, read PPTX, extract text from DOCX, convert document to markdown, or mentions .pdf/.docx/.pptx/.xlsx/.xls/.ppt/.doc files for reading or conversion. Triggers on phrases like "convert…
  • 100obsidian-file-intel
    Extract content from files (PDF, PPTX, XLSX, DOCX, CSV, JSON) and generate Obsidian-ready summaries. Use when processing a folder, summarizing documents, or running file intel. ファイル処理・要約。檔案處理・摘要。

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.