Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Feishu Docx Exporter · leemysw | DecimalAI

leemysw/feishu-docx

SkillScore100Maintained

Export, write, and manage Feishu/Lark cloud documents. Supports docx, sheets, bitable, wiki, WeChat article import/export, drive management, and browser-based export for public or browser-readable docs. Use this skill when you need to read, analyze, write, or manage content in a Feishu knowledge base.

Documentationby leemysw

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Knowledge Management pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/leemysw-feishu-docx/SKILL.md

decimalai skills pull leemysw-feishu-docx --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/leemysw-feishu-docx/SKILL.md https://app.decimal.ai/s/leemysw-feishu-docx/SKILL.md

Raw file: /s/leemysw-feishu-docx/SKILL.md ↗ · machine-readable summary at /s/leemysw-feishu-docx.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

Runs

all versions · 30d

Eval pass

100%

global

Used when offered

56%

used / offered · all versions

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

23 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 5d ago

Yes — +87% lift+20% tokens0% turns100% of whole cases pass with skill (23/23)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-09	✗→✓	▲ Improved	31%	0%
case-10	✗→✓	▲ Improved	-17%	0%
case-01	✗→✓	▲ Improved	-7%	0%
case-02	✗→✓	▲ Improved	-2%	0%
case-03	✗→✓	▲ Improved	45%	0%

Showing 5 of 23 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

leemysw / feishu-docx

.skills/feishu-docx/SKILL.md

View on GitHub ↗

License MITImported 8/3/2026Version v1

Skill contents

SKILL.md

Feishu Docx Exporter

Export Feishu/Lark cloud documents to Markdown for AI analysis, writing, and automation.

Setup (One-time)

bash
pip install feishu-docx
feishu-docx config set --app-id YOUR_APP_ID --app-secret YOUR_APP_SECRET

> Token auto-refreshes. No user interaction required.

Optional features are split into extras so the core install stays lightweight. Quote extras in zsh and similar shells.

Optional: Browser-Based Export

export-browser requires Playwright and a Chromium runtime:

bash
pip install 'feishu-docx[browser]'
playwright install chromium

Optional: PDF Export

PDF export requires the pdf extra:

bash
pip install 'feishu-docx[pdf]'

Export Documents

bash
feishu-docx export "<FEISHU_URL>" -o ./output

The exported Markdown file will be saved with the document's title as filename.

If the document is public or only readable in your current browser session, prefer:

bash
feishu-docx export-browser "<FEISHU_OR_LARK_URL>" -o ./output

Or reuse an existing Playwright session:

bash
feishu-docx export-browser "<FEISHU_OR_LARK_URL>" --storage-state ./storage_state.json

Supported Document Types

docx: Feishu cloud documents → Markdown with images
sheet: Spreadsheets → Markdown tables
bitable: Multidimensional tables → Markdown tables
wiki: Knowledge base nodes → Auto-resolved and exported
public/browser-readable docs: Browser-based export with local images and attachments

Command Reference

| Command | Description | |---------|-------------| | feishu-docx export <URL> | Export document to Markdown | | feishu-docx export <URL> --pdf | Export document to Markdown and PDF | | feishu-docx export-browser <URL> | Export in a real browser session with local assets | | feishu-docx export-wechat <URL> | Export WeChat article to Markdown | | feishu-docx create <TITLE> | Create new document | | feishu-docx create --url <URL> | Create document from WeChat article | | feishu-docx write <URL> | Append content to document | | feishu-docx update <URL> | Update specific block | | feishu-docx drive ls | List app or personal cloud-space files | | feishu-docx drive perm-show <TOKEN> | Show public permission | | feishu-docx drive perm-members <TOKEN> | List permission members | | feishu-docx drive clear | Clear files with double confirmation | | feishu-docx export-wiki-space <URL> | Batch export entire wiki space | | feishu-docx export-workspace-schema <ID> | Export bitable database schema | | feishu-docx auth | OAuth authorization | | feishu-docx config set | Set credentials | | feishu-docx config show | Show current config | | feishu-docx config clear | Clear token cache | | feishu-docx tui | Interactive TUI interface |

Examples

Export a wiki page

bash
feishu-docx export "https://xxx.feishu.cn/wiki/ABC123" -o ./docs

Export a document with custom filename

bash
feishu-docx export "https://xxx.feishu.cn/docx/XYZ789" -o ./docs -n meeting_notes

Export a public or browser-readable doc in a real browser session

bash
feishu-docx export-browser "https://xxx.larkoffice.com/wiki/ABC123" -o ./browser_docs

Export a document as PDF

bash
feishu-docx export "https://xxx.feishu.cn/docx/XYZ789" --pdf

Read content directly (recommended for AI Agent)

bash
# Output content to stdout instead of saving to file
feishu-docx export "https://xxx.feishu.cn/wiki/ABC123" --stdout
# or use short flag
feishu-docx export "https://xxx.feishu.cn/wiki/ABC123" -c

Export with Block IDs (for later updates)

bash
# Include block IDs as HTML comments in the Markdown output
feishu-docx export "https://xxx.feishu.cn/wiki/ABC123" --with-block-ids
# or use short flag
feishu-docx export "https://xxx.feishu.cn/wiki/ABC123" -b

Batch Export Entire Wiki Space

bash
# Export all documents in a wiki space (auto-extract space_id from URL)
feishu-docx export-wiki-space "https://xxx.feishu.cn/wiki/ABC123" -o ./wiki_backup

# Specify depth limit
feishu-docx export-wiki-space "https://xxx.feishu.cn/wiki/ABC123" -o ./docs --max-depth 3

# Export with Block IDs for later updates
feishu-docx export-wiki-space "https://xxx.feishu.cn/wiki/ABC123" -o ./docs -b

Export Database Schema

bash
# Export bitable/workspace database schema as Markdown
feishu-docx export-workspace-schema <workspace_id>

# Specify output file
feishu-docx export-workspace-schema <workspace_id> -o ./schema.md

Write Documents (CLI)

Create Document

bash
# Create empty document
feishu-docx create "我的笔记"

# Create with Markdown content
feishu-docx create "会议记录" -c "# 会议纪要\n\n- 议题一\n- 议题二"

# Create from Markdown file
feishu-docx create "周报" -f ./weekly_report.md

# Create in specific folder
feishu-docx create "笔记" --folder fldcnXXXXXX

# Create from a WeChat article URL
feishu-docx create --url "https://mp.weixin.qq.com/s/xxxxx"

如何获取 folder token:

在浏览器中打开目标文件夹
从 URL 中提取 token：https://xxx.feishu.cn/drive/folder/fldcnXXXXXX
fldcnXXXXXX 就是 folder token

Append Content to Existing Document

bash
# Append Markdown content
feishu-docx write "https://xxx.feishu.cn/docx/xxx" -c "## 新章节\n\n内容"

# Append from file
feishu-docx write "https://xxx.feishu.cn/docx/xxx" -f ./content.md

Manage Drive Files

bash
# List app cloud-space documents
feishu-docx drive ls --auth-mode tenant --type docx

# List personal cloud-space documents
feishu-docx drive ls --auth-mode oauth --type docx

# Show public permission
feishu-docx drive perm-show "https://xxx.feishu.cn/docx/ABC123"

# List permission members
feishu-docx drive perm-members "https://xxx.feishu.cn/docx/ABC123"

# Clear files with double confirmation
feishu-docx drive clear --type docx

Update Specific Block

bash
# Step 1: Export with Block IDs
feishu-docx export "https://xxx.feishu.cn/docx/xxx" -b -o ./

# Step 2: Find block ID from HTML comments
# <!-- block:blk123abc -->
# # Heading
# <!-- /block -->

# Step 3: Update the specific block
feishu-docx update "https://xxx.feishu.cn/docx/xxx" -b blk123abc -c "新内容"

> Tip for AI Agents: When you need to update a specific section: > 1. Export with -b to get block IDs > 2. Find the target block ID from HTML comments > 3. Use feishu-docx update with that block ID

Tips

Images and attachments auto-download to {doc_title}/ folder when local assets are available
Prefer export-browser for public share links or browser-readable docs
Use --stdout or -c for direct content output (recommended for agents)
Use -b to export with block IDs for later updates
Token auto-refreshes, no re-auth needed
For Lark (overseas): add --lark flag
tenant_access_token manages app cloud space, user_access_token manages personal cloud space

Want to see if leemysw/feishu-docx would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/24/2026, 18:51 UTC · 23 cases

Lift

+87%

100% vs 13%

Agent turns

1.0 vs 1.0

Tokens

+20%

2,559.3 vs 2,131.43

fail → pass · 20pass → pass · 3

Per-case results (23)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-09 I am using browser-based export for a session-restricted doc at https://company.feishu.cn/docx/SEC123 and need to reuse an existing session state file at `./storage_state.json`. How should I invoke the export?	fail→pass	10,019	2,054	-79%	1	1	0%	1,793	2,351	+31%	0	0	—
case-10 I want to create a new cloud document imported directly from a WeChat article at `https://mp.weixin.qq.com/s/sample_article_id`. What command accomplishes this in one step?	fail→pass	17,244	2,629	-85%	1	1	0%	2,946	2,456	-17%	0	0	—
case-01 I need to view the contents of a Feishu document at https://mycompany.feishu.cn/wiki/ABC123. Please output the document's raw Markdown straight to my screen rather than creating a saved file, and embed the section block IDs inside the text so I can programmatically edit specific blocks later.	fail→pass	14,322	4,014	-72%	1	1	0%	2,790	2,602	-7%	0	0	—
case-02 Please download all pages from our company wiki space starting at https://mycompany.feishu.cn/wiki/SPACE123 into a local folder named `./kb_backup`. Make sure to limit the export recursive depth to 3 levels deep.	fail→pass	13,029	2,667	-80%	1	1	0%	2,520	2,464	-2%	0	0	—
case-03 I have a local report file located at `./weekly_report.md` that I want added to an existing Lark doc. Can you take its content and attach it to the end of the document at https://mycompany.larkoffice.com/docx/DOC123?	fail→pass	9,330	3,147	-66%	1	1	0%	1,732	2,503	+45%	0	0	—
case-04 I have a local Markdown file `summary.md` on my laptop that I want to convert into a rendered PDF file locally without uploading it to any cloud services. How can I perform this conversion via CLI?	pass→pass	11,598	9,304	-20%	1	1	0%	1,928	3,659	+90%	0	0	—
case-05 I need to fetch and export documentation pages from an enterprise Confluence space at https://confluence.example.com/display/TEAM into local Markdown files. What command or script should I use?	pass→pass	13,563	7,036	-48%	1	1	0%	2,287	3,346	+46%	0	0	—
case-06 How can I programmatically replace heading text in a local Markdown document on disk using a Python script?	pass→pass	13,437	7,707	-43%	1	1	0%	2,447	3,443	+41%	0	0	—
case-07 I want to export a cloud document at https://company.feishu.cn/docx/XYZ789 directly to Markdown and generate a PDF output via CLI. What command should I execute?	fail→pass	10,875	2,565	-76%	1	1	0%	1,968	2,412	+23%	0	0	—
case-08 I have a public document link https://open.larkoffice.com/wiki/PUB123 that requires a real browser rendering session to export local images without app API credentials. What command should I use?	fail→pass	8,428	2,351	-72%	1	1	0%	1,460	2,452	+68%	0	0	—
case-11 I need to create a new document titled 'Project Roadmap' containing markdown text '# Overview' inside a specific folder with token `fldcn999888`. What CLI command should I run?	fail→pass	12,486	2,423	-81%	1	1	0%	1,971	2,453	+24%	0	0	—
case-12 How can I extract the database schema of a multidimensional bitable table with workspace ID `tbl_workspace_555` to a local Markdown file `./schema.md`?	fail→pass	16,110	2,445	-85%	1	1	0%	3,189	2,442	-23%	0	0	—
case-13 I want to list docx documents stored in the app's cloud space using tenant authentication mode. What command should I run?	fail→pass	13,472	1,723	-87%	1	1	0%	2,348	2,278	-3%	0	0	—
case-14 I need to list docx documents stored in my personal cloud space using user OAuth authorization mode. What command should I execute?	fail→pass	10,969	2,257	-79%	1	1	0%	1,533	2,347	+53%	0	0	—
case-15 I exported a document with block IDs and identified block `blk99999`. How do I update only this specific block in the document at https://company.feishu.cn/docx/DOC555 with text 'Updated Heading'?	fail→pass	10,110	2,093	-79%	1	1	0%	1,952	2,413	+24%	0	0	—
case-16 How do I set up my App ID `cli_a1b2c3d4e5` and App Secret `sec_x1y2z3` for CLI authentication so token refresh works automatically?	fail→pass	13,107	2,377	-82%	1	1	0%	2,255	2,431	+8%	0	0	—
case-17 What command inspects public sharing permissions for document token `docx_sample_token_123`?	fail→pass	5,867	1,883	-68%	1	1	0%	988	2,300	+133%	0	0	—
case-18 How do I view permission members associated with document token `docx_sample_token_123`?	fail→pass	10,554	3,139	-70%	1	1	0%	1,944	2,359	+21%	0	0	—
case-19 I need to clear all docx files from the app cloud space with confirmation. What CLI command performs this operation?	fail→pass	8,847	2,669	-70%	1	1	0%	1,474	2,419	+64%	0	0	—
case-20 What python package extras installation and setup steps are required to enable browser rendering export support for Lark documents?	fail→pass	16,255	3,580	-78%	1	1	0%	2,545	2,586	+2%	0	0	—
case-21 What pip installation command adds optional dependencies required specifically for PDF export functionality?	fail→pass	7,656	1,742	-77%	1	1	0%	1,150	2,225	+93%	0	0	—
case-22 I want to convert a public WeChat article at `https://mp.weixin.qq.com/s/article123` directly into Markdown without saving it as a cloud document first. What command does this?	fail→pass	9,677	2,375	-75%	1	1	0%	1,552	2,416	+56%	0	0	—
case-23 How can I batch export an entire wiki space from https://company.feishu.cn/wiki/SPACE777 into `./wiki_out` while preserving block IDs inside output Markdown files?	fail→pass	22,570	3,065	-86%	1	1	0%	4,251	2,507	-41%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +87 percentage points is the difference between those two pass rates over the 23 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 13%
Pass rate with the skill: 100%
Cases attempted: 23
Cases counted toward lift: 23
Cases that regressed: 0
Skill version measured: v1
Run completed: 8/24/2026, 18:51 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-09

Prompt

I am using browser-based export for a session-restricted doc at https://company.feishu.cn/docx/SEC123 and need to reuse an existing session state file at `./storage_state.json`. How should I invoke the export?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals2 of 4 complete

BenchmarkPass rate WITH skill on a verified run100

Live evalPass rate on real evaluated traces100%

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+57% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 2 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 4 days ago

Other measured skills in the registry, with their headline benchmark lift.