一键导入
verify
Verify claude-commit changes by driving the cco CLI end-to-end on real staged changes - launch/drive recipe and gotchas.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Verify claude-commit changes by driving the cco CLI end-to-end on real staged changes - launch/drive recipe and gotchas.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
SDK for building AI agents with Claude Code's capabilities. Programmatically interact with Claude to build autonomous agents that can understand codebases, edit files, and execute workflows. ALWAYS use when writing code importing "@anthropic-ai/claude-agent-sdk". Consult for debugging, best practices, or modifying @anthropic-ai/claude-agent-sdk, anthropic-ai/claude-agent-sdk, anthropic-ai claude-agent-sdk, anthropic ai claude agent sdk, claude-agent-sdk-typescript, claude agent sdk typescript.
Use Claude, Anthropic's AI assistant, right from your terminal. Claude can understand your codebase, edit files, run terminal commands, and handle entire workflows for you. ALWAYS use when writing code importing "@anthropic-ai/claude-code". Consult for debugging, best practices, or modifying @anthropic-ai/claude-code, anthropic-ai/claude-code, anthropic-ai claude-code, anthropic ai claude code, claude-code-2.1.88, claude code 2.1.88.
ALWAYS use when writing code importing "@opentui/core". Consult for debugging, best practices, or modifying @opentui/core, opentui/core, opentui core, opentui.
Prettier is an opinionated code formatter. ALWAYS use when writing code importing "prettier". Consult for debugging, best practices, or modifying prettier.
ALWAYS use when writing code importing "commander". Consult for debugging, best practices, or modifying commander, commander.js.
Generate AI agent skills from npm package documentation. ALWAYS use when writing code importing "skilld". Consult for debugging, best practices, or modifying skilld.
| name | verify |
| description | Verify claude-commit changes by driving the cco CLI end-to-end on real staged changes - launch/drive recipe and gotchas. |
The surface is the cco CLI (bin/cco.ts). Tests and typecheck are CI's job;
verification means running the CLI against real staged changes.
git add -Abun run bin/cco.ts --dry-run --no-spinner --verbose
claude binary.--model-summary bogus-model-xyz and expect a
clean error naming the model, exit code 1. Put the bogus model on
--model-summary (not --model-final) so the run fails before any paid
summary call.bun run bin/cco.ts --dry-run -i (OpenTUI; needs a TTY).allowApiKey is enabled.head - $? reports the last
command in the pipeline, not cco.claude binary (e.g. sonnet ->
claude-sonnet-5). To see the resolved model, context window, and cost, run
./node_modules/.bin/claude -p 'Reply OK' --model <m> --output-format json
and read the modelUsage key.