Skill planning worksheet: text to speech

Complete this before asking a coding agent to change files or make a live provider request.

1. Purpose and human owner

• One small outcome:
• Intended listener and use:
• Human who decides whether the result may be used:
• Original, public-domain, fictional, or authorized test text:

2. Command contract

• Command name:
• Required voice selector:
• Exactly-one text inputs:
• Required output path:
• Overwrite rule: refuse by default; --force only when deliberately supplied.
• Dry-run behavior:
• Expected output format:

3. Provider and voice boundary

• Reference provider: Speechify for this lab; current documentation checked on:
• Adapter boundary: what stays provider-specific inside the implementation?
• Approved local voice slug and provider voice ID:
• Why this voice is appropriate for this test:
• What the command must never choose automatically:

4. Secrets and cost boundary

• Local secret file location: .env (ignored)
• Committed example contains names only: .env.example
• Ignore rule checked:
• Spend limit or alert checked:
• Where the key must never appear: prompts, source, Git history, logs, screenshots, receipts, static sites, or browser bundles.
• If exposure is suspected: revoke or rotate the key, remove access where it was exposed, and record the incident without repeating the secret.

5. Acceptance tests

Record expected result and actual result for each:

Test	Expected result	Actual result
--dry-run with valid inputs	No synthesis; clear planned request summary
Missing --out	Refusal
Missing text	Refusal
Both text inputs	Refusal
Unknown/unapproved voice	Refusal
Missing credential	Refusal; no request attempted
Existing output without --force	Refusal
One short live test	Valid audio with a positive duration

6. Human review

• Output path and duration:
• Billable character count, if returned:
• Can a listener understand it without the transcript?
• Voice fit observation:
• Decision: approved for the stated use / revise / not for use
• Next human action:

---

Source: labs/build-a-safe-text-to-speech-skill/materials/skill-planning-worksheet.md