[ahmed-vapi]

Description

• Session management: document creating a session with squadId. Handoffs between squad members stay inside the session, so context carries over. Keep the separate-session pattern for independent assistants. Links to the Create session API reference.
• Streaming: both TypeScript samples now keep only deltas whose path ends in .content. With a squad, handoff tool call arguments also stream as deltas and otherwise show up in the reply text. Added a note explaining the filter.
• Email address reading: replace "speak slowly" and "pause briefly" prompt instructions with comma-separated parts, since the LLM only outputs text. Added a tip pointing to punctuation (per the prompting guide) and the voice speed setting.
• Prompting guide + downloadable reference: replace the "Before responding, silently verify" block with a direct condition-and-action rule. Most voice agents run on non-reasoning models with no hidden step for that check, so it can be skipped or spoken.

Edited with the docs edit pass skill against the current style guide.

Testing Steps

• Run the app locally using fern docs dev or navigate to preview deployment
• Ensure that the changed pages and code snippets work

🤖 Generated with Claude Code

[lightsage-app]

Lightsage docs evals

Result: passed
Staging docs: https://vapi-preview-01a0efb2-8930-74aa-85df-02ab1b13511e.docs.buildwithfern.com
Commit: f68c581

Average score: 94/100
Passed: 1/1

Eval ID	Status	Score	Model	Tools	Docs 404
08e24c18-82a0-45de-abdc-d237bd12bc0f	Pass	94	claude-code/global.anthropic.claude...	13	1

[github-actions]

🌿 Preview your docs: https://vapi-preview-01a0efab-d8df-776a-a029-31b71f736719.docs.buildwithfern.com

[github-actions]

🌿 Preview your docs: https://vapi-preview-01a0efb2-8930-74aa-85df-02ab1b13511e.docs.buildwithfern.com