MUST USE when managing persistent session state, avoiding context overflow, or enabling session handoffs. Implements the scratchpad pattern for solo developers working with AI agents.
Manages persistent session state using a scratchpad pattern to avoid context overflow. Use when context exceeds 60% capacity, working on multi-step tasks, or needing session handoffs between tools.
/plugin marketplace add Git-Fg/thecattoolkit/plugin install git-fg-thecattoolkit@Git-Fg/thecattoolkitThis skill inherits all available tools. When active, it can use any tool Claude has access to.
references/context-structure.mdtemplates/handoff.mdtemplates/scratchpad.mdworkflows/create-handoff.mdworkflows/initialize.mdworkflows/summarize-session.mdworkflows/update-scratchpad.mdAnalyze the user's request to determine the appropriate workflow:
Manage persistent session state independent of the LLM context window using the Scratchpad Pattern. Preserve decisions, track progress, and enable seamless session handoffs.
Use context-engineering when:
Don't use when:
Core Problem: AI models have limited context windows, but complex tasks require extensive context preservation across interactions.
Solution: Engineer context to maximize relevant information while minimizing token consumption through persistent storage in .claude/context/.
Use when: Starting new project or task requiring persistent state
Creates:
.claude/context/ directory structurescratchpad.md with project contexttodos.md for persistent task trackingcontext.log for session historycheckpoints/ directory for state snapshotsUse when: Recording decisions, errors, progress, or important context
Updates:
Use when: Consolidating session state before rotation or handoff
Produces:
Use when: Transferring work to new session or different tool
Generates:
.claude/context/
├── scratchpad.md # Current thinking and decisions
├── todos.md # Persistent task tracking
├── plan.md # Current implementation plan
├── context.log # Session context history
├── handoff.md # Session handoff summary
└── checkpoints/ # Critical state snapshots
├── 2025-01-05-feature-start.md
├── 2025-01-05-implementation-complete.md
└── 2025-01-05-testing-phase.md
Track these metrics:
Works seamlessly with:
For Solo Developers:
Workflow Improvements:
initialize workflow to set up context structureupdate-scratchpad workflow to record decisions and progresscreate-handoff when rotating sessionsFor detailed guidance, see Context Structure Reference.
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.