build-zoom-virtual-agent — Claude command: embed Zoom Virtual Agent
Walks an agent through embedding Zoom Virtual Agent on web, Android or iOS: codebase recon, plan, edits, verification.
~/.codex/prompts/build-zoom-virtual-agent.mdWhat build-zoom-virtual-agent does
Reach for this command when a repo has to embed or wrap Zoom Virtual Agent on web, Android or iOS. It starts by inspecting the codebase for existing web views, wrapper bridges, event handlers and the code that passes auth or context. Then it pins down the target platform and the agent launch path, checks the host app lifecycle constraints and confirms the required configuration values without printing secrets.
Before any edits the command asks you to state the plan: target platform and Virtual Agent workflow, the files to change, the launch path, event handling and what the host app owns, plus how the first integration will be verified. The edits themselves are narrow — embed or wrapper code at the right platform layer, the minimum launch, event and context-update logic for a working path. Wrapper concerns are kept apart from broader app state, and tokens stay out of logs and output.
Verification means re-reading the changed embed or bridge files, running a local build or tests where they exist, and checking the shape of the launch and message path. The command names remaining blockers outright: missing platform config, unsupported host behaviour, gaps in app setup. If wrapper lifecycle handling turns out unstable, it tells you to narrow the integration before expanding scope.
What it helps with
- embed Zoom Virtual Agent in a web view or native wrapper
- split agent launch and event handling across platform layers
- align the integration with the host app lifecycle
- keep tokens and secrets out of logs
- separate wrapper state from broader application state
Example prompts for the agent
- embed Zoom Virtual Agent in our web app
- add a Zoom Virtual Agent wrapper for Android
- check the launch path and event handling for Zoom Virtual Agent
How to install build-zoom-virtual-agent
Manually: put the files in place
~/.codex/prompts/build-zoom-virtual-agent.mdCall it with /prompts:build-zoom-virtual-agent. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.claude/commands/build-zoom-virtual-agent.mdCall it with /build-zoom-virtual-agent. For all projects use ~/.claude/commands/
Manually: put the files in place
.github/prompts/build-zoom-virtual-agent.prompt.mdManually: put the files in place
.gemini/commands/build-zoom-virtual-agent.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about build-zoom-virtual-agent
How do I install build-zoom-virtual-agent in Claude?
Place the command file at ~/.codex/prompts/build-zoom-virtual-agent.md.
What does build-zoom-virtual-agent do?
It runs a Zoom Virtual Agent integration on web, Android or iOS: codebase recon, a plan, embed edits and verification of the launch and event path.
What does the command check before making changes?
It looks for existing web views, bridges, event handlers and auth-passing code, identifies the target platform and launch path, and confirms host lifecycle constraints and required configuration values.
Source text
build-zoom-virtual-agent.md · MIT
--- description: Implement a Zoom Virtual Agent integration for web or mobile wrapper environments with the right lifecycle and event handling. --- # Build Zoom Virtual Agent Use this command when the repo needs to embed or wrap Zoom Virtual Agent on web, Android, or iOS. ## Preflight 1. Inspect the codebase for existing web views, wrapper bridges, event handlers, and auth or context-passing code. 2. Identify the target platform and Virtual Agent launch path. 3. Confirm the lifecycle constraints for the host app and any required native bridge behavior. 4. Check for the required configuration values without printing secrets. ## Plan Before making changes: - state the target platform and Virtual Agent workflow - list the files that will be changed - state the launch path, event handling path, and host-app responsibilities - state how the initial integration will be verified ## Commands 1. Add or correct the Virtual Agent embed or wrapper code at the appropriate platform layer. 2. Keep the implementation aligned with the host framework’s lifecycle and bridge patterns. 3. Add the minimum launch, event, and context update logic required for a working path. 4. Separate wrapper concerns from broader application state where possible. 5. Avoid leaking secrets or user tokens in logs or output. ## Verification 1. Re-read the embed or wrapper code and any supporting bridge files that changed. 2. Run local build or tests where available. 3. Verify the launch and message or event path shape. 4. State any remaining blocker such as missing platform config, unsupported host behavior, or app setup gaps. ## Summary ```text ## Result - Action: implemented or updated a Zoom Virtual Agent workflow - Status: success | partial | failed - Details: platform, files changed, launch path, verification run ``` ## Next Steps - Test the embedded agent in the intended runtime. - Add deeper host-app integration only after the base launch path works. - If wrapper lifecycle handling is unstable, narrow the integration before expanding scope.
© Zoom · MIT · openai/plugins