Token导航 LogoToken导航TokenDH.com
研究检索敏感数据github未标认证来源可访问clear审计提醒

opencode-expertOpenCode 专家

Agent Skill

opencode-expert 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

7,246

周安装

296

GitHub Stars

4

下载量

2,344
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:opencode-expert(OpenCode 专家)
来源仓库:https://github.com/s-hiraoku/synapse-a2a
仓库路径:skills/opencode-expert
安装命令:
npx skills add https://github.com/s-hiraoku/synapse-a2a --skill opencode-expert
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/s-hiraoku/synapse-a2a --skill opencode-expert

简介

opencode-expert 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。

  • 它可结合来源仓库、安装命令和原始 README 继续核验具体用法。
  • 安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。
  • 可通过 npx skills add 命令从 GitHub 仓库安装。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

OpenCode Expert

Comprehensive guide for OpenCode - the open-source AI coding agent.

Quick Reference

TaskCommand/Action
Start TUIopencode
Continue sessionopencode -c or opencode --continue
Run non-interactiveopencode run "message"
Start headless serveropencode serve
Web interfaceopencode web
Switch agentTab key
File search@ then type filename
Undo changes/undo
Redo changes/redo
Share conversation/share
Initialize project/init
Configure API keys/connect

Agents

OpenCode has two primary agents, switchable with Tab:

Build Agent (Default)

  • Full access to all tools
  • Can edit files, run commands, make changes
  • Use for active development work

Plan Agent

  • Read-only mode for analysis and exploration
  • File edits and bash commands require approval (ask mode)
  • Use for reviewing implementation strategy before changes

Subagents

Invoke subagents with @ mention:

  • @general - Complex research and multi-step tasks with full tool access
  • @explore - Fast read-only codebase searches and code questions

Keyboard Shortcuts

KeyAction
TabSwitch between Build and Plan agents
@Fuzzy file search
Ctrl+CCancel current operation

Slash Commands

CommandDescription
/connectConfigure LLM provider API keys
/initInitialize project and create AGENTS.md
/undoRevert recent changes (stackable)
/redoRestore undone changes
/shareCreate shareable conversation link

CLI Commands

Interactive Mode

# Start TUI in current directory
opencode

# Start in specific project
opencode /path/to/project

# Continue previous session
opencode --continue
opencode -c

# Use specific session
opencode --session <id>
opencode -s <id>

# Start with specific model
opencode --model <model>
opencode -m <model>

# Start with specific agent
opencode --agent build
opencode --agent plan

Non-Interactive Mode

# Run single prompt
opencode run "Fix the bug in auth.py"

# Run with file attachment
opencode run "Review this" --file README.md
opencode run "Review this" -f README.md

# Run and share result
opencode run "Explain this code" --share

# Run specific command
opencode run --command "npm test"

Server Mode

# Start headless API server
opencode serve

# With custom port
opencode serve --port 3000

# With mDNS discovery
opencode serve --mdns

# Web interface
opencode web

Management Commands

# Authentication
opencode auth login
opencode auth list
opencode auth logout

# Models
opencode models              # List available models
opencode models --refresh    # Refresh model cache

# Agents
opencode agent create        # Create custom agent
opencode agent list          # List agents

# MCP servers
opencode mcp add <server>
opencode mcp list
opencode mcp auth
opencode mcp debug

# Sessions
opencode session list
opencode session list -n 10  # Last 10 sessions

# Statistics
opencode stats
opencode stats --days 7
opencode stats --tools
opencode stats --models

# Export/Import
opencode export <sessionID>
opencode import <file>

# Upgrade
opencode upgrade
opencode upgrade <version>

Built-in Tools

OpenCode provides these tools to agents:

ToolDescription
readRead file contents with optional line ranges
writeCreate new files or overwrite existing
editModify files using exact string replacements
patchApply patch files
grepSearch file contents with regex
globFind files by pattern (**/*.js)
listList files and directories
bashExecute shell commands
lspLanguage Server Protocol integration (experimental)
todowriteCreate/update task lists
todoreadRead current todo state
skillLoad skill files into conversation
webfetchFetch and read web pages
questionAsk user clarifying questions

Skills

File Locations

Skills are discovered from (in order):

  1. .agents/skills/<name>/SKILL.md (project, shared with Codex)
  2. .claude/skills/<name>/SKILL.md (Claude-compatible project)
  3. ~/.config/opencode/skills/<name>/SKILL.md (global)
  4. ~/.claude/skills/<name>/SKILL.md (Claude-compatible global)

Skill Format

---
name: my-skill
description: Description of what this skill does and when to use it.
---

# Skill Content

Instructions and documentation here...

Name Rules

  • Lowercase alphanumeric with single hyphens
  • Cannot start/end with hyphens
  • No consecutive hyphens (--)
  • Must match directory name

Configuration

Config Locations

  • Project: .opencode/ or opencode.json
  • Global: ~/.config/opencode/

Environment Variables

VariableDescription
OPENCODE_CONFIGCustom config path
OPENCODE_PERMISSIONPermission mode
OPENCODE_SERVER_PASSWORDServer auth password
OPENCODE_DISABLE_AUTOUPDATEDisable auto-updates
OPENCODE_EXPERIMENTALEnable experimental features
OPENCODE_ENABLE_EXAEnable Exa search

Custom Agents

Create agents via opencode agent create or define in:

  • opencode.json
  • ~/.config/opencode/agents/
  • .opencode/agents/

Agent config options: description, temperature, model, tool permissions, system prompts.

Best Practices

Use Plan Mode First

Switch to Plan agent (Tab) to review implementation strategy before making changes.

File References

Use @filename to fuzzy search and reference files in prompts.

Undo Stack

Use /undo liberally - changes are stackable and reversible with /redo.

Session Management

Use --continue or -c to resume previous sessions and maintain context.

References

For detailed documentation, see:

  • references/cli.md - Complete CLI reference
  • references/tools.md - Tool details and permissions
  • references/agents.md - Agent configuration guide

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenCode

26.48%
按下载量换算621

Claude Code

19.86%
按下载量换算466

Gemini CLI

17.77%
按下载量换算417

Cursor

13.37%
按下载量换算313

Codex

8.39%
按下载量换算197

Antigravity

3.47%
按下载量换算81

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。

来源信息

继续浏览同类 Skills