Initial commit

2025-11-30 08:56:13 +08:00
commit 966e739b00
9 changed files with 1805 additions and 0 deletions
--- a/skills/tmux-task-runner/EXAMPLES.md
+++ b/skills/tmux-task-runner/EXAMPLES.md
@@ -0,0 +1,325 @@
+# Tmux Task Runner - Examples
+
+This document provides practical examples of using the tmux-task-runner skill.
+
+## Quick Start
+
+### Example 1: Running a Build Process
+
+```bash
+# Start a build in a tmux session
+./run.sh run build "npm run build"
+
+# Output:
+# ✓ Task started in tmux session
+#
+# Session: task-build-1729519263
+# Log file: $LOG_DIR/task-build-1729519263.log   (LOG_DIR defaults to /tmp)
+# Status file: $STATUS_DIR/task-build-1729519263.status
+# Workdir: /Users/you/project
+#
+# Monitoring commands:
+#   tail -f $LOG_DIR/task-build-1729519263.log
+#   tmux attach-session -t task-build-1729519263
+#   ./run.sh tail task-build-1729519263
+#   tmux kill-session -t task-build-1729519263
+```
+
+Monitor the build:
+```bash
+tail -f "$LOG_DIR"/task-build-1729519263.log
+```
+
+### Example 2: Running Tests
+
+```bash
+# Run pytest with verbose output and CI environment
+./run.sh run test --env NODE_ENV=ci "pytest tests/ --verbose --cov"
+
+# Or with npm/yarn (preserve exit status with coverage)
+./run.sh run test --notify "npm test -- --coverage"
+```
+
+Check test status:
+```bash
+./run.sh check task-test-1729519263
+# Or review the recorded metadata:
+./run.sh status task-test-1729519263
+```
+
+### Example 3: Starting a Development Server
+
+```bash
+# Start a Python HTTP server
+./run.sh run server "python -m http.server 8000"
+
+# Start a Node.js dev server
+./run.sh run server --workdir ./apps/frontend "npm run dev"
+
+# Start a Django dev server
+./run.sh run server "python manage.py runserver"
+```
+
+Attach to the server session to see real-time requests:
+```bash
+./run.sh attach task-server-1729519263
+# Press Ctrl+b then d to detach
+```
+
+### Example 4: Running Deployment Scripts
+
+```bash
+# Deploy to production
+./run.sh run deploy --notify "./deploy.sh production"
+
+# Or with a deployment tool
+./run.sh run deploy "terraform apply -auto-approve"
+```
+
+Monitor deployment progress:
+```bash
+# Poll the last 80 lines every 5 seconds
+./run.sh tail task-deploy-1729519263 --interval 5 --lines 80
+```
+
+## Advanced Usage
+
+### Example 5: Managing Multiple Concurrent Tasks
+
+```bash
+# Start multiple tasks
+./run.sh run build "npm run build"
+./run.sh run test "npm test"
+./run.sh run lint "npm run lint"
+
+# List all active tasks
+./run.sh list
+
+# Output:
+# Active task sessions:
+# task-build-1729519263: 1 windows (created Tue Oct 21 10:30:00 2025)
+# task-test-1729519264: 1 windows (created Tue Oct 21 10:30:05 2025)
+# task-lint-1729519265: 1 windows (created Tue Oct 21 10:30:10 2025)
+```
+
+### Example 6: Long-Running Data Processing
+
+```bash
+# Process large dataset
+./run.sh run process "python scripts/process_data.py --input large_dataset.csv"
+
+# Machine learning training
+./run.sh run train "python train_model.py --epochs 100"
+```
+
+Check progress periodically:
+```bash
+# Every 30 seconds, check the last 20 lines
+./run.sh tail task-train-1729519263 --interval 30 --lines 20
+```
+
+### Example 7: Database Migrations
+
+```bash
+# Run database migrations
+./run.sh run migrate "alembic upgrade head"
+
+# Or with Django
+./run.sh run migrate "python manage.py migrate"
+```
+
+### Example 8: Cleanup Operations
+
+```bash
+# Check all running tasks
+./run.sh list
+
+# Kill a specific task
+./run.sh kill task-build-1729519263
+
+# Kill all task sessions
+./run.sh kill all
+```
+
+## Real-World Scenarios
+
+### Scenario 1: CI/CD Pipeline Simulation
+
+```bash
+# Run a complete CI/CD pipeline locally
+./run.sh run ci "bash -c '
+  echo \"=== Installing dependencies ===\" &&
+  npm install &&
+  echo \"=== Running linter ===\" &&
+  npm run lint &&
+  echo \"=== Running tests ===\" &&
+  npm test &&
+  echo \"=== Building application ===\" &&
+  npm run build &&
+  echo \"=== CI Pipeline Complete ===\"
+'"
+
+# Monitor the pipeline
+tail -f "$LOG_DIR"/task-ci-*.log
+```
+
+### Scenario 2: Monitoring API Load Testing
+
+```bash
+# Start API load test
+./run.sh run loadtest "ab -n 10000 -c 100 http://localhost:8000/api/endpoint"
+
+# Or with a more advanced tool
+./run.sh run loadtest "artillery run load-test-config.yml"
+
+# Attach to watch real-time metrics
+./run.sh attach task-loadtest-1729519263
+```
+
+### Scenario 3: File Processing Pipeline
+
+```bash
+# Process video files
+./run.sh run encode "ffmpeg -i input.mp4 -c:v libx264 -preset slow output.mp4"
+
+# Batch image optimization
+./run.sh run optimize "find ./images -name '*.jpg' -exec mogrify -quality 85 {} \;"
+```
+
+### Scenario 4: Development Workflow
+
+```bash
+# Terminal 1: Start backend server
+./run.sh run backend "python manage.py runserver"
+
+# Terminal 2: Start frontend dev server
+./run.sh run frontend "npm run dev"
+
+# Terminal 3: Watch tests
+./run.sh run test-watch "npm run test:watch"
+
+# Terminal 4: Monitor all logs
+./run.sh list
+# Then tail specific logs as needed
+```
+
+## Integration with Claude Code
+
+When working with Claude Code, you can request task execution like this:
+
+**User:** "Run the test suite in the background so I can continue working"
+
+**Claude:** I'll run your test suite in a tmux session using the tmux-task-runner skill.
+
+```bash
+./skills/tmux-task-runner/run.sh run test "npm test -- --coverage"
+```
+
+Your tests are now running in session `task-test-1729519263`.
+
+Monitor progress:
+```bash
+./skills/tmux-task-runner/run.sh tail task-test-1729519263 --lines 80
+# Or follow the log directly:
+tail -f "$LOG_DIR"/task-test-1729519263.log
+```
+
+I'll check the initial output:
+```bash
+tmux capture-pane -t task-test-1729519263 -p -S -20
+```
+
+The tests are running. You can continue working, and I'll check back on the results in a few moments.
+
+---
+
+**User:** "Check on the test results"
+
+**Claude:** Let me check the test session status:
+
+```bash
+./skills/tmux-task-runner/run.sh check task-test-1729519263
+```
+
+[Shows output with test results]
+
+The tests have completed successfully! All 127 tests passed with 92% coverage.
+
+## Tips and Tricks
+
+### Tip 1: Creating Task Aliases
+
+Add to your `~/.bashrc` or `~/.zshrc`:
+
+```bash
+alias tmux-task='/path/to/skills/tmux-task-runner/run.sh'
+
+# Now you can use:
+tmux-task run build "npm run build"
+tmux-task list
+```
+
+### Tip 2: Monitoring Multiple Sessions
+
+```bash
+# Create a monitoring dashboard using tmux itself
+tmux new-session -s monitoring \; \
+  send-keys "tail -f \"$LOG_DIR\"/task-build-*.log" C-m \; \
+  split-window -v \; \
+  send-keys "tail -f \"$LOG_DIR\"/task-test-*.log" C-m \; \
+  split-window -h \; \
+  send-keys "watch -n 5 './run.sh list'" C-m
+```
+
+### Tip 3: Saving Output
+
+```bash
+# Save complete session output to file
+tmux capture-pane -t task-build-1729519263 -p -S - > build-output.txt
+
+# Or just use the log file
+cp "$LOG_DIR"/task-build-1729519263.log ./build-output.txt
+```
+
+### Tip 4: Setting Up Notifications
+
+```bash
+# Built-in flag handles macOS/Linux automatically
+./run.sh run build --notify "npm run build"
+```
+
+## Troubleshooting
+
+### Issue: Can't find session
+
+```bash
+# List all sessions to find the correct name
+tmux list-sessions
+
+# Or check available logs
+ls -lt "$LOG_DIR"/task-*.log
+```
+
+### Issue: Session closed unexpectedly
+
+```bash
+# Check the log file for errors
+cat "$LOG_DIR"/task-build-1729519263.log
+
+# Look for error messages or exit codes
+tail -50 "$LOG_DIR"/task-build-1729519263.log
+```
+
+### Issue: Want to keep session after task completes
+
+Sessions exit automatically once the command completes, but the log and status files remain for review (`./run.sh status <session>`).
+
+### Issue: Too many old sessions
+
+```bash
+# Kill all active task sessions
+./run.sh kill all
+
+# Clean up old log files (older than 7 days)
+find "$LOG_DIR" -name "task-*.log" -mtime +7 -delete
+```