zoom-cobrowse-sdk · openai — Codex skill: what it does, how to install
Plan and debug Zoom Cobrowse browser sessions, including participant roles, privacy controls, session setup, and reconnection.
npx skills add openai/plugins --skill cobrowse-sdk -a codexWhat zoom-cobrowse-sdk does
Use this skill once a support flow has settled on browser co-browsing with Zoom Cobrowse. It helps define the customer page and agent console roles, decide how a PIN or link reaches the other participant, and check authentication before UI work begins. If the product choice is still between Cobrowse and other Zoom products, start with start or plan-zoom-product.
The workflow starts with tokens, allowed origins, environment variables, and SDK loading. It then covers masking, blocked elements, and limits on remote assistance. For the base lifecycle, initialize the SDK, start or join a session, handle events, reconnect, and end the session.
Add annotations, multi-tab persistence, custom PINs, or remote assistance after the core session is stable. During debugging, inspect browser support, CORS/CSP, token generation, third-party cookies, and SDK event errors. This skill guides integration work; it does not help choose a Zoom product.
What it helps with
- map customer and agent roles in a Cobrowse session
- check authentication and SDK loading before UI work
- set privacy masking and remote-assistance boundaries
- trace connection, CORS/CSP, or SDK event failures
Example prompts for the agent
- Help me set up Zoom Cobrowse: define the customer and agent roles, then outline session startup and joining.
- Check what my project needs for SDK loading, authentication, and handing off a session PIN.
- Debug this Cobrowse failure by checking browser support, CORS/CSP, token generation, third-party cookies, and SDK events.
How to install zoom-cobrowse-sdk
1Skills CLI (Vercel)
npx skills add openai/plugins --skill cobrowse-sdk -a codex2Manually: put the files in place
.agents/skills/cobrowse-sdk/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add openai/plugins --skill cobrowse-sdk -a claude-code2Manually: put the files in place
~/.claude/skills/cobrowse-sdk/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/cobrowse-sdk/
Manually: put the files in place
.cursor/skills/cobrowse-sdk/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install openai/plugins cobrowse-sdk2Manually: put the files in place
.github/skills/cobrowse-sdk/SKILL.mdPlatform CLI
gemini skills install https://github.com/openai/plugins --path plugins/zoom/skills/cobrowse-sdkManually: put the files in place
.devin/skills/cobrowse-sdk/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: Codex skills: how to install, where they live, how to call them
Questions about zoom-cobrowse-sdk
How do I install zoom-cobrowse-sdk in Codex?
Run npx skills add openai/plugins --skill cobrowse-sdk -a codex. The skill file is at .agents/skills/cobrowse-sdk/SKILL.md.
What does zoom-cobrowse-sdk do?
It guides the setup of Zoom Cobrowse browser sessions, including their lifecycle and privacy controls, and helps investigate SDK or connection errors.
What should I check before building the Cobrowse UI?
Check tokens, allowed origins, environment variables, and SDK loading. Define participant roles and how the PIN or link will be handed off as well.
Source text
SKILL.md · MIT
--- name: zoom-cobrowse-sdk description: Use when using Cobrowse. --- # Zoom Cobrowse SDK Use this skill after the support workflow is clearly a browser co-browsing experience. If the user is still choosing between Zoom Contact Center, Virtual Agent, Video SDK, or Cobrowse, route through `start` or `plan-zoom-product` first. ## Workflow 1. Confirm the role model: customer page, agent console, session initiation, and how the PIN or link is handed off. 2. Check auth and session setup before UI work: tokens, allowed origins, environment variables, and SDK loading. 3. Design privacy controls early: masking rules, blocked elements, remote-assist permission boundaries, and audit requirements. 4. Implement the minimal lifecycle first: initialize, start or join, subscribe to events, reconnect, and end. 5. Add advanced capabilities only after the base session is stable: annotations, multi-tab persistence, custom PINs, and remote assist. 6. When debugging, isolate browser support, CORS/CSP, token generation, third-party cookie behavior, and SDK event errors. ## References - Full preserved guide: [references/full-guide.md](references/full-guide.md) - API reference: [references/api-reference.md](references/api-reference.md) - Session lifecycle: [concepts/session-lifecycle.md](concepts/session-lifecycle.md) - Privacy masking: [examples/privacy-masking.md](examples/privacy-masking.md) - Remote assist: [examples/remote-assist.md](examples/remote-assist.md) - Common issues: [troubleshooting/common-issues.md](troubleshooting/common-issues.md)
© Zoom · MIT · openai/plugins
Similar skills
SKILL.md skills catalog →zoom-cobrowse-sdk by other authors comes first.