Skip to main content
reopt Handbook
reopt Handbook
Claude Code Command MasterLabels and command selectionHow to read a detail guideComplete command listTerminal CLI Commands and Flags

Task guides

Core command basicsStarting and resuming workSkills and Custom CommandsAutomation and Web CommandsOperations and Safety CommandsCustom commands and skillsMCP connections and commands

Conversation and context

Cmd. /add-dirCmd. /backgroundCmd. /compactCmd. /contextCmd. /copyCmd. /exportCmd. /forkCmd. /importCmd. /initCmd. /memoryCmd. /recapCmd. /renameCmd. /resumeCmd. /rewindCmd. claude purge

Planning and execution

Cmd. /autofix-prCmd. /batchCmd. /desktopCmd. /goalCmd. /loopCmd. /planCmd. /remote-envCmd. /scheduleCmd. /tasksCmd. /teleportCmd. /workflows

Review and verification

Cmd. /code-reviewCmd. /runCmd. /run-skill-generatorCmd. /security-reviewCmd. /simplifyCmd. /verify

Models and settings

Cmd. /claude-apiCmd. /configCmd. /effortCmd. /fastCmd. /modelCmd. /output-styleCmd. /update-config

Permissions and integrations

Cmd. /agentsCmd. /hooksCmd. /ideCmd. /mcpCmd. /permissionsCmd. /pluginCmd. /reload-pluginsCmd. /reload-skillsCmd. /sandboxCmd. /skillsCmd. /web-setup

Status and diagnostics

Cmd. /debugCmd. /diffCmd. /doctorCmd. /fewer-permission-promptsCmd. /release-notesCmd. /team-onboarding

Display and input

Cmd. /colorCmd. /focusCmd. /keybindingsCmd. /powerupCmd. /statuslineCmd. /terminal-setupCmd. /themeCmd. /tui

Account and usage

Cmd. /upgradeCmd. /usageCmd. /usage-credits

Legacy

Cmd. /costCmd. /pr-commentsCmd. /reviewCmd. /statsCmd. /ultrareviewCmd. /vimCmd. /approved-toolsCmd. /commandsCmd. /extra-usageCmd. /settingsCmd. /tagCmd. /todos

Verification and history

Current verification scopeHandbook updatesRelease coverage historyEarlier verification recordsCommand change history
Handbook›Claude Code Command Master›Cmd. /claude-api
한국어KOEnglishEN

Cmd. /claude-api

Claude API reference, migration, cost optimization, and evaluation workflows

Written and reviewed by Eric Han · Co-founder & CTO, reopt

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|api-version|managed-agents-onboard|managed-agents-secure|prompt-audit|cost-optimize|build-eval|hillclimb|preserved-thinking-migration]

Availability

FieldValue
Input surfaceRunning CLI prompt
KindBundled skill
StatusCurrent reference
Aliases—
Documented minimumNot specified
Reviewed against2026-10-23 · Claude Code 2.1.308

Subcommands have separate version requirements; build-eval and hillclimb are available in v2.1.259+, and preserved-thinking-migration in v2.1.282+. From v2.1.285 it cannot be run from Remote Control clients. managed-agents-secure requires v2.1.304+; api-version requires v2.1.307+.

Examples

/claude-api api-version
/claude-api managed-agents-secure
SubcommandUseDocumented minimum
migrateUpdate API code for a newer model.Not specified in Commands
upgradeMigrate the Python Anthropic SDK from 0.x to 1.x.v2.1.236
managed-agents-onboardCreate a Managed Agent with guided setup.Not specified in Commands
prompt-auditReview outdated model instructions and propose a diff.v2.1.221
cost-optimizeMeasure API spend and apply one measurable optimization at a time.v2.1.247
build-evalBuild an evaluation set for a Claude-powered app.v2.1.259
hillclimbImprove an app against an existing evaluation.v2.1.259
preserved-thinking-migrationFind edits to earlier turns, the system prompt, or the tool list that invalidate preserved thinking blocks, measure the reasoning each one drops, and propose fixes one at a time.v2.1.282
/claude-api build-eval
/claude-api hillclimb

Choose the evaluation target and success criteria first. Review code changes and run the evaluation after each change. All eight subcommands above are now in the official signature; the earlier note that cost-optimize was changelog-only is obsolete. From v2.1.290, managed-agents-onboard also takes a URL, setting up the Managed Agents pattern that page describes as ant apply files, or a Console quickstart name such as deep-researcher, building that template with the ant CLI.

Effects and persistence

Loads API guidance for the project's language or performs the selected subtask. Review file diffs from migration and optimization work.

  • /claude-api api-version (v2.1.307+) helps migrate code to API version 2026-10-21; distinguish it from model migration (migrate) and SDK migration (upgrade).
  • /claude-api managed-agents-secure (v2.1.304+) compares agent configuration with Anthropic's security guide without modifying settings.

Changelog.

Cautions

Subcommands have separate version requirements; build-eval and hillclimb are available in v2.1.259+. From v2.1.285 the skill cannot be run from Remote Control clients.

Official sources

  • Commands
  • Changelog
  • Skills

Complete command list · Verification scope

Related docs

Cmd. /fewer-permission-prompts

Propose a read-only allowlist from transcripts (skill).

Cmd. /run

Launch and drive the app to verify a change (skill).

Verification Report

Harness Engineering · Link, consistency, source, and static validation report for the Harness Engineering handbook.

Ch3. Evaluation Framework

LLMOps and AgentOps in Production · Connect offline benchmarks with online operating signals

Cmd. /verify

Build, run, and observe the app to verify a change (skill).

Cmd. /config

Open settings or apply direct key/value settings (/settings).

On this page

PurposeAvailabilityExamplesEffects and persistenceCautionsOfficial sources