AI Tools: Give Your Podio AI Real Power to Act

Until now, using AI in Podio through ProcFu meant asking questions and getting answers. Useful, but limited. The AI could tell you what to do — it just couldn't do it.

That changes with AI Tools.

What Are AI Tools?

AI Tools let your AI calls in ProcFu take action directly in Podio. Instead of just generating text, the AI can now create items, update fields, assign tasks, post comments, and attach files — all on its own, as part of its response.

Think of it as the difference between asking a colleague for advice and asking them to handle it. Same conversation, very different outcome.

Built-in Tools: Ready to Go

We've built six tools that cover the most common Podio write operations:

• Create — Create a new item in any Podio app
• Update — Update fields on an existing item
• Delete — Delete an item
• Task — Create and assign a task
• Comment — Add a comment to an item or task
• File — Attach a text file to an item

Enabling them is a one-liner. Add builtin_tools to your existing AI call:

podio_ai_ask_general(
    prompt = "Review this deal and create follow-up tasks for the sales team",
    item_id = 12345,
    builtin_tools = "task,comment"
)

Or go all in with builtin_tools = "all".

The AI is smart about it too — it automatically looks up your app structure so it knows exactly which fields exist and what to put where. No manual field mapping needed.

Custom Tools: This Is Where It Gets Powerful

Built-in tools handle common Podio operations. Custom tools blow the doors open.

A custom tool is a ProcScript block with three simple functions — name(), usage(), and run(). You describe what the tool does, and the AI decides when to call it. The key insight: inside run(), you have access to ProcFu's entire library of 300+ built-in ProcScript functions. That means your AI isn't limited to Podio — it can reach out to virtually any system.

What Can Custom Tools Actually Do?

Here's the short list of what's available inside a custom tool:

• Query databases — run SQL against MySQL databases with mysql_query
• Call any API — hit any REST endpoint with remote_curl or api_request_send
• Send emails — fire off notifications with email_send_smtp
• Work with Google Drive — list, create, move, and delete files with google_drive_* functions
• Read and write Google Sheets — query, append, and update spreadsheet data with sheets_*
• Manage Notion pages and databases — create pages, query tables, update rows with notion_* functions
• Transfer files via FTP — upload or download files with podio_files_to_ftp and podio_files_from_ftp
• Parse and create spreadsheets — work with Excel and CSV files using podio_file_xls_* and podio_file_csv_*
• Convert documents — transform between markdown, HTML, and PDF formats
• Search and filter Podio data — deep queries across apps, views, and spaces
• Store and retrieve variables — persist data across calls with var_get and var_set
• Trigger workflows — kick off GlobiFlow/PWA flows with pwa_trigger_flow_on_item

This is what makes custom tools fundamentally different from what most AI integrations offer. You're not limited to a handful of pre-built connectors. If ProcScript can do it — and ProcScript can do a lot — your AI can do it.

A Quick Example

Say you want the AI to be able to look up customer data and log interactions to a Google Sheet:

// Tool: lookup_customer
function name() {
    return "lookup_customer"
}

function usage() {
    return [
        "description" => "Look up a customer by email in Podio and return their details",
        "params" => [
            "email" => "required - The customer's email address"
        ]
    ]
}

function run($args) {
    $items = podio_item_filter(app_id=YOUR_APP_ID, filters=[
        ["field" => "email", "value" => $args.email]
    ])
    return $items
}

// Tool: log_to_sheet
function name() {
    return "log_interaction"
}

function usage() {
    return [
        "description" => "Log a customer interaction to the tracking spreadsheet",
        "params" => [
            "customer_name" => "required - Customer name",
            "action" => "required - What action was taken",
            "notes" => "Any additional notes"
        ]
    ]
}

function run($args) {
    sheets_append(
        spreadsheet_id = "YOUR_SHEET_ID",
        range = "Interactions!A:D",
        values = [[datetime_convert("now", "Y-m-d H:i"), $args.customer_name, $args.action, $args.notes]]
    )
    return "Logged successfully"
}

Now combine them:

podio_ai_ask_general(
    prompt = "Find the customer john@example.com, update their status to Active, and log this interaction",
    custom_tools = "lookup_customer,log_interaction",
    builtin_tools = "update"
)

The AI looks up the customer, updates the Podio item, and logs the interaction to Google Sheets — all from a single natural language instruction. Three systems, one prompt.

Real-World Scenarios

Because custom tools can reach any system ProcScript can talk to, the use cases go far beyond basic Podio automation:

• "Process this invoice" — AI reads the item, calls a custom tool that queries your MySQL database for the vendor's payment terms, creates a task with the due date, and sends a confirmation email via SMTP.

• "Sync this project to Notion" — AI reads the Podio item, calls a custom tool that creates a Notion page with the project details, and comments back on the Podio item with the Notion link.

• "Generate a weekly report" — AI uses a custom tool to query a Google Sheet for this week's metrics, another tool to filter Podio items by status, composes a summary, and emails it to the team.

• "Onboard this new client" — AI creates items across multiple Podio apps, sets up a Google Drive folder structure, adds rows to a tracking spreadsheet, sends a welcome email, and triggers a GlobiFlow onboarding sequence.

You can combine up to 10 custom tools with any of the built-in tools per call. The AI decides which tools to use, in what order, based on the prompt.

Works With the In-Podio AI Assistant Too

If you use aiNots — ProcFu's AI assistant that responds to @mentions and task assignments directly inside Podio — you can enable tools there as well. Just head to Account > AI Assistant and check the boxes for built-in tools or add your custom tool names. No code changes needed.

Getting Started

1. If you're already using podio_ai_ask_general or podio_ai_ask_trained, just add builtin_tools to your call. Start with a specific tool like "task" or "comment" before opening up to "all".
2. For custom tools, create a new ProcScript block with the three required functions, then reference it via custom_tools.
3. For the in-Podio assistant, configure tools in Account > AI Assistant.

All tool actions execute under your ProcFu account's Podio authentication and fire normal Podio webhook events — so your existing flows and integrations continue to work as expected.

Wrapping Up

Built-in tools get you started with common Podio operations in seconds. But custom tools are the real story here. With access to ProcFu's full library of 300+ functions — spanning databases, APIs, email, Google Drive, Sheets, Notion, FTP, file conversion, and more — you can give your AI the ability to operate across your entire tech stack.

The AI reasons. The tools act. And you define exactly what's possible.