From context-engine
Advanced context window management techniques. Guided compaction, partial compaction, side questions, checkpoints, session discipline. Use to master context management.
npx claudepluginhub artmin96/forge-studio --plugin context-engineThis skill uses the workspace's default tool permissions.
Guides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.
Migrates code, prompts, and API calls from Claude Sonnet 4.0/4.5 or Opus 4.1 to Opus 4.5, updating model strings on Anthropic, AWS, GCP, Azure platforms.
Analyzes BMad project state from catalog CSV, configs, artifacts, and query to recommend next skills or answer questions. Useful for help requests, 'what next', or starting BMad.
Context is your most precious resource. These techniques protect it.
Don't let auto-compaction decide what to keep. Guide it:
/compact preserve the auth refactoring plan and test results
Add to CLAUDE.md for persistent guidance:
When compacting, always preserve: current plan, test results, key architectural decisions.
Compress only part of the conversation:
/rewind — opens checkpoint menuAsk a quick question that doesn't enter conversation history:
/btw what's the syntax for Laravel's whereHas?
Answer comes back without polluting your working context.
Every Claude action creates a checkpoint. Use this aggressively:
/rewind → restore./rewind to open the menu.Include files instantly without waiting for Claude to read them:
@src/auth/middleware.php refactor this to use the new auth service
@src/api/ show me the API structure
Faster than waiting for Claude to Read the file. Paths can be relative or absolute.
Finish task → commit → /clear → new task
/clear and write a better promptclaude -n "auth-refactor" # Start named session
/rename auth-refactor # Rename during session
claude --resume auth-refactor # Resume by name later
At 60-70% context, Claude still has clear recall of the full conversation. Compacting here produces BETTER summaries than waiting until 90% when quality has already degraded. Watch your status line.
When you need to read many files for research:
Use a subagent to investigate src/auth/ and report what authentication patterns are used
The subagent reads 20 files. Only a 200-token summary returns to your session.