A selection of curated tools provided by the Autohive team

Actions

• Clock — Tool to get current time and convert between timezones.

Word document automation integration using python-docx with markdown-first content creation. AI agents should use markdown syntax for most content creation - headings (#), paragraphs, lists (- or 1.), formatting (**bold**, *italic*), tables, blockquotes (>), and code blocks. Only use non-markdown actions for images, page breaks, or when you need direct table creation with structured data arrays.

Actions

• Add Image — Insert an image into the document from uploaded files. Requires the current document file and image file in the 'files' parameter. The first image file found will be used. Supports PNG, JPG, JPEG, GIF, BMP, and WebP formats. Images cannot be added via markdown, so use this action for any image insertion. Specify width and/or height in inches to control sizing, or omit both for original size.
• Add Markdown Content — Parse and add markdown content to an existing Word document. Requires the current document file in the 'files' parameter. Use this to append additional content to documents created with create_document. All markdown syntax is supported: headings (#), formatting (**bold**, *italic*), lists (- bullets, 1. numbers), tables, blockquotes (>), code blocks (```), etc. Content is appended to the end of the document. Returns the updated document file.
• Add Page Break — Insert a page break into the document to force content onto the next page. Requires the current document file in the 'files' parameter. Page breaks cannot be represented in standard markdown, so use this action when you need to control page layout. Returns the updated document file.
• Add Structured Table — Create a table in the document using structured data arrays. Requires the current document file in the 'files' parameter. Use this as an alternative to markdown tables when you have structured data in array format or need precise control over table dimensions. The data parameter should be a 2D array where each sub-array represents a row. First row is typically headers. Returns the updated document file.
• Analyze Document Structure — Scan a document template and return all modifiable elements with their content and position indices. Use this to analyze templates before filling them. Automatically detects fillable elements like placeholders ({{field}}, [field], 'data here', etc.) and empty cells. Returns paragraph and table structure for precise template filling.
• Create Word Document — Create a new Word document with markdown content. This is the primary action for document creation. Use standard markdown syntax: # for headings (# H1, ## H2, ### H3, etc.), plain text for paragraphs, - for bullet lists, 1. for numbered lists, **bold**, *italic*, `code`, > for blockquotes, and markdown tables (| col1 | col2 |). The markdown is automatically converted to properly formatted Word elements with appropriate styles. Returns a unique document_id for use with other actions and the complete document file for stateless operation.
• Find and Replace Text — Search for exact text phrases and replace them throughout the document with advanced safety features and spacing control. CRITICAL SAFETY RULE: Always provide sufficient context in your find phrases to avoid unintended replacements. SAFE EXAMPLES: 'Date: ___' → 'Date: January 15, 2025', 'Project Name: placeholder text' → 'Project Name: Q4 Integration'. UNSAFE EXAMPLES: 'date' → 'January 15' (will corrupt 'the update date'). SPACING CONTROL: Use remove_paragraph=true to eliminate gaps when deleting instruction paragraphs like '(Note: Delete this instruction)' → ' '. Use remove_paragraph=false (default) to preserve intentional document spacing. System prevents multiple unintended matches by showing all match locations when replace_all=false.
• Save Document — Save a document and return it as a base64-encoded file for download. The document is saved to memory and returned as streaming content - no physical file is created on the server. This is typically called automatically by other actions, but can be used standalone if needed.
• Update Elements by Position — Modify specific document elements by their position indices. Use get_document_elements first to find element positions. Supports updating paragraphs by index and table cells by table index + coordinates. Preserves original formatting while updating content.

Create and edit PDF files from text, images, or other PDFs

Actions

• Create and Edit PDF — Perform PDF tasks like creating, merging, rotating, extracting pages, or watermarking. Accepts files as input and returns resulting PDFs.