Add audio pronunciation (TTS) for words and hints

[llinsss]

Problem

The entire hint and story system (agent/hint_generator.py, agent/story_mode.py) is text-only. The target users are children learning to read — many can't decode the hint text itself, which undermines the product's core purpose.

Scope

• Integrate a TTS provider (e.g. Amazon Polly, since the project already uses AWS/boto3 for Bedrock) to generate audio for: the target word, each hint level, and generated stories.
• Add an endpoint (or extend existing ones) to return an audio URL/stream alongside text, e.g. POST /api/v1/hint response gains an audio_url field.
• Cache generated audio (don't regenerate on every request for the same word/hint) — a local file cache keyed by content hash is sufficient.
• Fall back gracefully (text-only) if the TTS provider is unavailable or unconfigured, mirroring the existing Bedrock fallback pattern.

Acceptance criteria

• /api/v1/hint and /api/v1/story responses include an audio reference when TTS is configured.
• Requesting the same word/hint twice reuses cached audio (verified by mocking the TTS client and asserting it's called once).
• With TTS unconfigured/erroring, endpoints still return valid text-only responses (test mirrors the existing test_bedrock_story_fallback_on_error pattern).

[Lspnjr1]

Hello maintainers,

I’d like to take on this issue. I’ve gone through the issue details and believe I can implement the required changes effectively. I’ll ensure the solution is properly tested, follows the project’s conventions, and submit a pull request as soon as it’s ready.

Please assign this issue to me if it’s available. Thank you!

[Lspnjr1]

Hello maintainers,

I’d like to take on this issue. I’ve gone through the issue details and believe I can implement the required changes effectively. I’ll ensure the solution is properly tested, follows the project’s conventions, and submit a pull request as soon as it’s ready.

Please assign this issue to me if it’s available. Thank you!

[Lspnjr1]

Hello maintainers,

I’d like to take on this issue. I’ve gone through the issue details and believe I can implement the required changes effectively. I’ll ensure the solution is properly tested, follows the project’s conventions, and submit a pull request as soon as it’s ready.

Please assign this issue to me if it’s available. Thank you!

[Lspnjr1]

Hello maintainers,

I’d like to take on this issue. I’ve gone through the issue details and believe I can implement the required changes effectively. I’ll ensure the solution is properly tested, follows the project’s conventions, and submit a pull request as soon as it’s ready.

Please assign this issue to me if it’s available. Thank you!

[grantfox-oss]

🦊 GrantFox — @Chisom92 has been assigned to this issue as part of the Official Campaign | FWC26 campaign!

Next steps:

1. Open a Pull Request referencing this issue (e.g., Closes #4)
2. Your PR will be reviewed by the DogStark maintainers

Good luck! Track your progress on GrantFox.