[Agent] ElevenLabs TTS tool

[Guikingone]

Q	A
Bug fix?	no
New feature?	yes
Docs?	yes
Issues	None
License	MIT

Hi 👋🏻

This PR aims to introduce a new native tool to allows the TTS pipeline from the agent, the HTTP call is performed against ElevenLabs API (only for TTS currently).

Once received, the file is stored using the $path property, the unit test is performed against the fixture file (not sure about this one but it works).

[Copilot]

Pull Request Overview

This PR introduces a new native tool for ElevenLabs Text-to-Speech (TTS) functionality within the Symfony AI Agent package. The tool enables agents to convert text to speech by making HTTP calls to the ElevenLabs API and storing the generated audio files locally.

Key changes:

• Adds ElevenLabs TTS tool with configurable voice, model, and output path
• Includes comprehensive test coverage using mock HTTP client
• Provides example usage and documentation updates

Reviewed Changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
src/agent/src/Toolbox/Tool/ElevenLabs.php	Core TTS tool implementation with ElevenLabs API integration
src/agent/tests/Toolbox/Tool/ElevenLabsTest.php	Unit tests for the ElevenLabs tool functionality
src/agent/composer.json	Adds symfony/filesystem dependency for file operations
examples/misc/text-to-speech.php	Working example demonstrating tool usage
src/agent/doc/index.rst	Documentation updates referencing the new example
examples/.env	Environment variable for ElevenLabs API key

Comments suppressed due to low confidence (1)

src/agent/tests/Toolbox/Tool/ElevenLabsTest.php:38

• The test doesn't verify that the file was actually created or that the 'path' key contains the expected file path. Consider adding assertions to validate the file creation and path structure.

        $this->assertCount(2, $result);

[chr-hertel]

ElevenLabs 🥳

Great stuff - thanks for having a look at that platform! So there are multiple things in here - and we should split it:

instead of focussing on the tool, let's bring in a Platform bridge first please. basically a Symfony\AI\Platform\Bridge\ElevenLabs and text input, binary result

you're twisting my mind here, but you're right - we should out of the box enable agents to use platforms as a tool as well - not only agents.
and maybe even 3, for example in the demo project there is some kind of pre processing with whisper (speech2text) and this could be also post processing text2speech - this should be standardized somehow ...

a lot of stuff in those 165 lines of code 😆

are you up for starting with the platform bridge?

[Guikingone]

are you up for starting with the platform bridge?

Yes, of course.

I'll take a look at the pre-processing part 👍🏻