Agent Skills: Using Quip

Using Quip

Interact with Quip (axon.quip.com) via the Quip Automation REST API. Supports reading, creating, editing, searching, and organizing documents and folders.

When to Use

• User provides an axon.quip.com URL and wants to read, edit, or discuss the document
• User asks to create a new Quip document or folder
• User wants to search Quip for documents by keyword
• User wants to append content to an existing Quip document
• User asks about Quip folder structure or wants to organize documents

When NOT to Use

• Real-time collaboration or commenting on specific selected text (use the Quip UI)
• Uploading large binary attachments (use the Blobs API directly)
• Anything requiring OAuth flows — only Bearer token auth is supported here

Configuration

# Required: set in shell profile or export before invoking Claude
export QUIP_ACCESS_TOKEN="your-personal-access-token"

| Setting | Value | | ------------------ | ---------------------------------------------- | | API base URL | https://platform.quip.com/1/ | | Human-facing URL | https://axon.quip.com/{thread_id} | | Auth header | Authorization: Bearer $QUIP_ACCESS_TOKEN | | POST body encoding | application/x-www-form-urlencoded (NOT JSON) |

CRITICAL: The API base is platform.quip.com, NOT axon.quip.com. The tenant domain is only used for human-facing links.

NEVER echo or log $QUIP_ACCESS_TOKEN in output.

Extracting Thread IDs from URLs

Given https://axon.quip.com/XYZ123AbC, the thread ID is XYZ123AbC (the path segment after the domain).

# Extract thread ID from URL
THREAD_ID=$(echo "https://axon.quip.com/XYZ123AbC" | sed 's|.*/||')

Content Format

Quip documents use HTML. Always use format=html when creating or editing content. Markdown loses fidelity on ingestion.

| HTML Tag | Purpose | | ----------------------- | ------------- | | <h1>, <h2>, <h3> | Headings | | <p> | Paragraph | | <b>, <i> | Bold, italic | | <ul><li>...</li></ul> | Bulleted list | | <ol><li>...</li></ul> | Numbered list | | <code> | Inline code | | <pre> | Code block |

Edit Location Enum

When editing documents, location controls where content is inserted:

| Value | Intent | Requires section_id? | | ----- | ----------------------------- | ---------------------- | | 0 | Append to end of document | No | | 1 | Prepend to start of document | No | | 2 | After the specified section | Yes | | 3 | Before the specified section | Yes | | 4 | Replace the specified section | Yes | | 5 | Delete the specified section | Yes |

Extracting section IDs: Fetch the document HTML, then find data-section-id attributes:

curl -s -H "Authorization: Bearer $QUIP_ACCESS_TOKEN" \
  "https://platform.quip.com/1/threads/THREAD_ID" \
  | python3 -c "
import json,sys
from html.parser import HTMLParser

class SectionParser(HTMLParser):
    def handle_starttag(self, tag, attrs):
        for name, val in attrs:
            if name == 'data-section-id':
                print(f'{tag}: {val}')

data = json.load(sys.stdin)
SectionParser().feed(data['html'])
"

Quick Start

1. Verify Credentials

curl -s -H "Authorization: Bearer $QUIP_ACCESS_TOKEN" \
  "https://platform.quip.com/1/users/current" | python3 -m json.tool

2. Read a Document

# Given URL: https://axon.quip.com/ABC123
curl -s -H "Authorization: Bearer $QUIP_ACCESS_TOKEN" \
  "https://platform.quip.com/1/threads/ABC123" | python3 -m json.tool

3. Create a Document

curl -s -X POST \
  -H "Authorization: Bearer $QUIP_ACCESS_TOKEN" \
  -H "Content-Type: application/x-www-form-urlencoded" \
  --data-urlencode "title=My Document" \
  --data-urlencode "content=<h1>Title</h1><p>Body text.</p>" \
  --data-urlencode "format=html" \
  --data-urlencode "type=document" \
  "https://platform.quip.com/1/threads/new-document" | python3 -m json.tool

4. Append to a Document

curl -s -X POST \
  -H "Authorization: Bearer $QUIP_ACCESS_TOKEN" \
  -H "Content-Type: application/x-www-form-urlencoded" \
  --data-urlencode "thread_id=ABC123" \
  --data-urlencode "content=<p>New paragraph appended.</p>" \
  --data-urlencode "location=0" \
  --data-urlencode "format=html" \
  "https://platform.quip.com/1/threads/edit-document" | python3 -m json.tool

5. Search Documents

curl -s -H "Authorization: Bearer $QUIP_ACCESS_TOKEN" \
  "https://platform.quip.com/1/threads/search?query=quarterly+review&count=10" \
  | python3 -m json.tool

Workflows

Reading a Document

1. Extract thread ID from the axon.quip.com URL
2. GET threads/{thread_id} — returns thread.html (full content) and metadata
3. Parse HTML for structure; use data-section-id attributes if editing is needed
4. Present content in a readable format to the user

Creating a Document

1. Compose content as HTML
2. POST threads/new-document with title, content, format=html, type=document
3. Optionally specify member_ids to share with users/folders immediately
4. Return the document URL: https://axon.quip.com/{thread.thread.id}

Editing a Document

1. Fetch the document to get current HTML and section IDs
2. Determine location enum (0=append, 4=replace section, etc.)
3. For targeted edits: extract section_id from data-section-id attributes
4. POST threads/edit-document with thread_id, content, location, optionally section_id

Searching for Documents

1. GET threads/search?query={terms}&count={n} — returns array of thread objects
2. Present titles and URLs to user
3. User can select a result → fetch full document with GET threads/{id}

Navigating Folder Structure

1. GET users/current → note private_folder_id and shared_folder_ids
2. GET folders/{folder_id} → children array contains thread_link and folder_link items
3. Recurse into subfolders as needed
4. Use folder IDs when creating documents in a specific location (member_ids)

Organizing Documents

1. Create new folder: POST folders/new with title, optionally parent_id and color
2. Move document to folder: POST folders/add-members with folder_id and member_ids
3. Remove document from folder: POST folders/remove-members

Posting Messages

1. Get thread ID (document or chat thread)
2. POST messages/new with thread_id and content
3. Optionally use author_id for named attribution

Error Handling

| HTTP Status | Meaning | Action | | ----------- | ---------------------------- | -------------------------------------------- | | 400 | Bad request / invalid params | Check param names and encoding | | 401 | Unauthorized | Verify $QUIP_ACCESS_TOKEN is set and valid | | 403 | Forbidden | User lacks access to this thread/folder | | 404 | Not found | Check thread/folder ID is correct | | 429 | Rate limited | Wait and retry with exponential backoff | | 500 | Server error | Retry once; escalate if persistent |

Validation Checklist

After any write operation:

• [ ] Response HTTP status is 200
• [ ] Response body contains expected fields (thread.id, folder.id, etc.)
• [ ] Document URL resolves: https://axon.quip.com/{id}
• [ ] For edits: fetch document again and verify content change is present
• [ ] Token was NOT echoed in any output

Reference Files

For complete endpoint specs:

Read ~/.claude/skills/using-quip/references/API-REFERENCE.md

Use when: You need exact parameter names, response schemas, or pagination patterns.

For complete working examples:

Read ~/.claude/skills/using-quip/references/EXAMPLES.md

Use when: You need a full curl recipe for a specific workflow.