What problem does it solve?
This Skill eliminates the tedious manual work of tracking complex, multi-phase workflow states and dependencies. It provides a persistent, queryable database for all workflow data, drastically reducing the time and token cost associated with parsing files for analytics and state management.
Core Features & Use Cases
- Persistent State Management: Stores workflow states, tasks, quality gates, and checkpoints in AgentDB (DuckDB) for reliable, long-term tracking.
- Advanced Workflow Analytics: Enables complex queries for task dependencies, critical paths, historical metrics, and state transition analysis, offering deep insights into project progress.
- Context Checkpoint Management: Automatically saves and restores AI context at token thresholds (e.g., 100K tokens), allowing for faster recovery and efficient session management.
- Token Efficiency: Reduces token usage by up to 92% for complex queries compared to manual file parsing, optimizing AI operational costs.
- Healthcare Compliance: Supports HIPAA, FDA 21 CFR Part 11, and IRB compliance with immutable audit trails and detailed Protected Health Information (PHI) access tracking.
- Use Case: An AI agent is managing a large software development project. This skill allows it to instantly answer questions like "What tasks are currently blocking progress?" or "Show me the historical trend of test coverage over the last 30 days," without re-parsing all project files, thus accelerating decision-making and project oversight.
Quick Start
To analyze the current workflow state and task dependencies, use the following command:
python .claude/skills/agentdb-state-manager/scripts/query_state.py --dependencies
