/adapter
You are orchestrating the sigcli agent team to add a new browser adapter. The user's request: $ARGUMENTS
$ npx -y skills add sigcli/sigcli --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/adapter
Context preview
What this command does when you run it.
You are orchestrating the sigcli agent team to add a new browser adapter. The user's request: $ARGUMENTS
Command definition
adapter.md/adapter — Add New Browser Adapter
You are orchestrating the sigcli agent team to add a new browser adapter. The user's request: $ARGUMENTS
Context
Browser adapters implement `IBrowserAdapter`, `IBrowserSession`, and `IBrowserPage` from `src/core/interfaces/browser-adapter.ts`. The reference implementation is `PlaywrightAdapter` in `src/browser/adapters/playwright.adapter.ts` (three-class pattern: Adapter → Session → Page).
Workflow
Phase 1: Design
Spawn the **architect** agent with this prompt: > Design a new browser adapter for the sigcli project: $ARGUMENTS > > You MUST read these files first: > 1. `src/core/interfaces/browser-adapter.ts` — the IBrowserAdapter/IBrowserSession/IBrowserPage contract > 2. `src/browser/adapters/playwright.adapter.ts` — the reference implementation (three-class pattern) > 3. `src/core/types.ts` — BrowserLaunchOptions, Cookie types > > Design the adapter following the three-class pattern. Map every IBrowserPage method to the target library's equivalent. Identify which library methods need adaptation. Plan the file structure, exports, and tests.
Present the plan to the user. **Wait for approval.**
Phase 2: Implement
Spawn the **dev** agent with the approved plan: > Implement the new browser adapter: > [Include architect's plan] > > Create `src/browser/adapters/<name>.adapter.ts` with three classes (Adapter, Session, Page). Lazy-import the browser library. Throw BrowserLaunchError on import failure. Export the Adapter class. Add it to `src/index.ts`. Run npm run build.
Phase 3: Test
Spawn the **tester** agent: > Write unit tests for the new browser adapter at `src/browser/adapters/<name>.adapter.ts`. > [Describe what was implemented] > > Create tests in `tests/unit/browser/`. Test: adapter has correct name property, launch method exists, Session and Page wrap the library correctly. Mock the underlying library. Run npm test.
Phase 4: Review
Spawn the **reviewer** agent: > Review the new browser adapter implementation. > [Describe what was built] > > Check: three-class pattern, lazy import, BrowserLaunchError on failure, all IBrowserPage methods mapped, exported in index.ts, tests exist. Run tsc --noEmit and npm test.
Phase 5: Report
Summarize: adapter created, files added, tests passing, review status.
Read more
/adapter — Add New Browser Adapter
You are orchestrating the sigcli agent team to add a new browser adapter. The user's request: $ARGUMENTS
Context
Browser adapters implement `IBrowserAdapter`, `IBrowserSession`, and `IBrowserPage` from `src/core/interfaces/browser-adapter.ts`. The reference implementation is `PlaywrightAdapter` in `src/browser/adapters/playwright.adapter.ts` (three-class pattern: Adapter → Session → Page).
Workflow
Phase 1: Design
Spawn the **architect** agent with this prompt: > Design a new browser adapter for the sigcli project: $ARGUMENTS > > You MUST read these files first: > 1. `src/core/interfaces/browser-adapter.ts` — the IBrowserAdapter/IBrowserSession/IBrowserPage contract > 2. `src/browser/adapters/playwright.adapter.ts` — the reference implementation (three-class pattern) > 3. `src/core/types.ts` — BrowserLaunchOptions, Cookie types > > Design the adapter following the three-class pattern. Map every IBrowserPage method to the target library's equivalent. Identify which library methods need adaptation. Plan the file structure, exports, and tests.
Present the plan to the user. **Wait for approval.**
Phase 2: Implement
Spawn the **dev** agent with the approved plan: > Implement the new browser adapter: > [Include architect's plan] > > Create `src/browser/adapters/<name>.adapter.ts` with three classes (Adapter, Session, Page). Lazy-import the browser library. Throw BrowserLaunchError on import failure. Export the Adapter class. Add it to `src/index.ts`. Run npm run build.
Phase 3: Test
Spawn the **tester** agent: > Write unit tests for the new browser adapter at `src/browser/adapters/<name>.adapter.ts`. > [Describe what was implemented] > > Create tests in `tests/unit/browser/`. Test: adapter has correct name property, launch method exists, Session and Page wrap the library correctly. Mock the underlying library. Run npm test.
Phase 4: Review
Spawn the **reviewer** agent: > Review the new browser adapter implementation. > [Describe what was built] > > Check: three-class pattern, lazy import, BrowserLaunchError on failure, all IBrowserPage methods mapped, exported in index.ts, tests exist. Run tsc --noEmit and npm test.
Phase 5: Report
Summarize: adapter created, files added, tests passing, review status.
Sign in your way. AI works on your behalf. AI agents need access to your work systems — Jira, wikis, calendars, internal APIs. But passing credentials through shell history, environment variables, and agent context windows is a security nightmare.

