Install any skill in seconds. Free to start, no credit card required.
Get Started Free →You are an expert in Cline, the autonomous AI coding agent for VS Code that can read files, write code, run terminal commands, and use the browser — with human-in-the-loop approval at each step. You help developers use Cline for complex multi-file refactoring, feature implementation, debugging, and codebase exploration where the AI plans and executes while the developer reviews and approves.
.claude/skills/terminalskills-cline/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -55% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 63% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 31% | 0% |
You are an expert in Cline, the autonomous AI coding agent for VS Code that can read files, write code, run terminal commands, and use the browser — with human-in-the-loop approval at each step. You help developers use Cline for complex multi-file refactoring, feature implementation, debugging, and codebase exploration where the AI plans and executes while the developer reviews and approves.
markdown## How Cline Works 1. You describe a task in natural language 2. Cline creates a plan (visible to you) 3. For each step, Cline: - Reads relevant files (auto-approved) - Proposes file changes (you approve/reject each) - Runs terminal commands (you approve first) - Opens browser for testing (you approve) 4. You can steer the agent mid-task ## Example: Add Authentication to an Existing App Prompt: "Add Clerk authentication to this Next.js app. Protect all /dashboard routes. Add a sign-in page. Show the user's name in the header." Cline's steps (each requires your approval): 1. Reads package.json, layout.tsx, middleware.ts 2. Runs: npm install @clerk/nextjs 3. Creates: src/middleware.ts with Clerk middleware 4. Modifies: src/app/layout.tsx to add ClerkProvider 5. Creates: src/app/sign-in/[[...sign-in]]/page.tsx 6. Modifies: src/components/Header.tsx to show UserButton 7. Runs: npm run build (to verify no errors) 8. Opens browser to test sign-in flow
json// VS Code settings for Cline { "cline.apiProvider": "anthropic", "cline.apiModelId": "claude-sonnet-4-20250514", "cline.customInstructions": "Follow the project conventions in .cursor/rules/. Use TypeScript strict mode. Prefer server components. Always run tests after changes.", "cline.autoApproveReadOnly": true, // Don't ask to read files "cline.maxFileLineCount": 500 // Warn before reading huge files }
markdown## Multi-File Refactoring "Migrate all API routes from Express to Hono. Keep the same route structure and middleware. Update tests to use the Hono test client." → Cline reads all route files, creates a migration plan, edits each file, updates imports, runs tests after each change. ## Feature Implementation "Add a Stripe webhook handler that processes subscription.created, subscription.updated, and subscription.deleted events. Update the user's plan in the database. Add proper signature verification." → Cline scaffolds the handler, adds event processing, creates DB migration, writes tests, and verifies the build. ## Debugging "Users report that the search API returns stale results. Find the caching layer, identify why it's not invalidating properly, and fix it." → Cline reads the search route, finds the Redis cache, identifies the TTL issue, proposes a fix with cache invalidation on write. ## Documentation "Generate API documentation for all endpoints in src/server/routers/. Create a markdown file for each router with endpoint descriptions, input/output schemas, and example requests." → Cline reads each router, extracts Zod schemas, and generates docs.
markdown## Cline supports MCP (Model Context Protocol) servers Add custom tools via MCP: - Database query tool → Cline can check DB schema/data - Deployment tool → Cline can deploy to staging - Testing tool → Cline can run specific test suites - Monitoring tool → Cline can check error rates Configure in Cline settings → MCP Servers → Add server URL
markdown# VS Code Extension Search "Cline" in VS Code Extensions marketplace # Or: code --install-extension saoudrizwan.claude-dev # Set API key in Cline settings # Supports: Anthropic, OpenAI, Google, AWS Bedrock, OpenRouter, local models
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 19,066 | 2,772 | -85% | 1 | 1 | 0% | 3,666 | 1,650 | -55% | 0 | 0 | — |
case-02 | fail→pass | 15,243 | 2,980 | -80% | 1 | 1 | 0% | 2,838 | 1,655 | -42% | 0 | 0 | — |
case-03 | pass→pass | 5,370 | 3,416 | -36% | 1 | 1 | 0% | 992 | 1,613 | +63% | 0 | 0 | — |
case-04 | pass→pass | 6,432 | 3,066 | -52% | 1 | 1 | 0% | 1,145 | 1,498 | +31% | 0 | 0 | — |
case-05 | pass→pass | 4,247 | 3,784 | -11% | 1 | 1 | 0% | 707 | 1,667 | +136% | 0 | 0 | — |
case-06 | fail→pass | 11,114 | 9,249 | -17% | 1 | 1 | 0% | 1,929 | 2,754 | +43% | 0 | 0 | — |
case-07 | pass→pass | 3,182 | 3,337 | +5% | 1 | 1 | 0% | 479 | 1,712 | +257% | 0 | 0 | — |
case-08 | pass→pass | 5,980 | 4,657 | -22% | 1 | 1 | 0% | 1,012 | 1,847 | +83% | 0 | 0 | — |
case-09 | pass→pass | 8,819 | 8,798 | -0% | 1 | 1 | 0% | 1,515 | 2,634 | +74% | 0 | 0 | — |
case-10 | pass→pass | 3,689 | 3,738 | +1% | 1 | 1 | 0% | 666 | 1,709 | +157% | 0 | 0 | — |
case-11 | pass→pass | 14,286 | 11,598 | -19% | 1 | 1 | 0% | 2,401 | 3,169 | +32% | 0 | 0 | — |
case-12 | pass→pass | 13,705 | 14,038 | +2% | 1 | 1 | 0% | 2,250 | 3,393 | +51% | 0 | 0 | — |
case-13 | pass→pass | 7,572 | 7,559 | -0% | 1 | 1 | 0% | 1,201 | 2,415 | +101% | 0 | 0 | — |
case-14 | pass→pass | 5,715 | 3,209 | -44% | 1 | 1 | 0% | 877 | 1,569 | +79% | 0 | 0 | — |
case-15 | pass→pass | 7,057 | 5,191 | -26% | 1 | 1 | 0% | 1,327 | 1,973 | +49% | 0 | 0 | — |
case-16 | pass→pass | 10,821 | 8,253 | -24% | 1 | 1 | 0% | 1,622 | 2,451 | +51% | 0 | 0 | — |
case-17 | pass→pass | 12,630 | 10,208 | -19% | 1 | 1 | 0% | 2,125 | 2,739 | +29% | 0 | 0 | — |
case-18 | pass→pass | 14,571 | 11,845 | -19% | 1 | 1 | 0% | 1,618 | 2,905 | +80% | 0 | 0 | — |
case-19 | fail→fail | 9,008 | 4,667 | -48% | 1 | 1 | 0% | 1,732 | 1,916 | +11% | 0 | 0 | — |
case-20 | pass→pass | 11,384 | 9,992 | -12% | 1 | 1 | 0% | 1,841 | 2,796 | +52% | 0 | 0 | — |
case-21 | pass→pass | 7,908 | 5,919 | -25% | 1 | 1 | 0% | 1,559 | 2,331 | +50% | 0 | 0 | — |
case-22 | pass→pass | 11,928 | 2,455 | -79% | 1 | 1 | 0% | 2,178 | 1,532 | -30% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.