Cmd. /claude-api
Claude API reference, migration, cost optimization, and evaluation workflows
Key takeaways
- Claude API and Managed Agents support skill.
- Check the input surface and availability before using the examples below.
Purpose
Claude API and Managed Agents support skill.
/claude-api [migrate|upgrade|managed-agents-onboard|prompt-audit|cost-optimize|build-eval|hillclimb]Availability
| Field | Value |
|---|---|
| Input surface | Running CLI prompt |
| Kind | Bundled skill |
| Status | Current reference |
| Aliases | — |
| Documented minimum | Not specified |
| Reviewed against | 2026-09-22 · Claude Code 2.1.278 |
Subcommands have separate version requirements; build-eval and hillclimb are available in v2.1.259+.
Examples
/claude-api prompt-audit| Subcommand | Use | Documented minimum |
|---|---|---|
migrate | Update API code for a newer model. | Not specified in Commands |
upgrade | Migrate the Python Anthropic SDK from 0.x to 1.x. | v2.1.236 |
managed-agents-onboard | Create a Managed Agent with guided setup. | Not specified in Commands |
prompt-audit | Review outdated model instructions and propose a diff. | v2.1.221 |
cost-optimize | Measure API spend and apply one measurable optimization at a time. | v2.1.247 |
build-eval | Build an evaluation set for a Claude-powered app. | v2.1.259 |
hillclimb | Improve an app against an existing evaluation. | v2.1.259 |
/claude-api build-eval
/claude-api hillclimbChoose the evaluation target and success criteria first. Review code changes and run the evaluation after each change. All seven subcommands above are now in the official signature; the earlier note that cost-optimize was changelog-only is obsolete.
Effects and persistence
Loads API guidance for the project's language or performs the selected subtask. Review file diffs from migration and optimization work.
Cautions
Subcommands have separate version requirements; build-eval and hillclimb are available in v2.1.259+.