Workflow Optimizer

If you're running complex Claude Code workflows with multiple subagents, tool chains, and model delegation, this gives you forensic level visibility into what's actually happening. It pulls eleven different datasets from the Agent Monitor API, from orchestration DAGs to tool transition matrices to error propagation patterns, then produces ranked optimization recommendations with estimated savings. You'd use this when your workflows feel inefficient but you're not sure where the waste is. The model delegation analysis alone is worth it if you're burning Opus tokens on tasks that Haiku could handle. It's basically a profiler for agentic workflows, assuming you're already using Agent Monitor to capture the telemetry.

GitHub stars: 455
last updated: Jun 11, 2026

View on GitHub

Install

$ npx -y skills add hoangsonww/claude-code-agent-monitor --skill workflow-optimizer --agent claude-code

Installs into .claude/skills of the current project.

Files

Workflow Optimizer

Analyze Claude Code workflows using the Agent Monitor's workflow intelligence engine.

Input

The user provides: $ARGUMENTS

Options: "analyze", a session ID for single-session analysis, or a focus: "tools", "subagents", "cost", "errors".

Data Sources

Endpoint	Returns
`GET /api/sessions?limit=100`	Session list with metadata
`GET /api/workflows/{sessionId}`	11 workflow datasets (see below)
`GET /api/analytics`	Tool usage top 20, event types, agent types
`GET /api/pricing`	Model pricing rules for cost comparison