collaborating-with-codex
Skill1 starsChanged 8 months agoMITView on GitHub
[由 collaborating-hub 路由] Codex CLI 后端实现。直接使用请通过 /collab 或 /codex 触发 collaborating-hub。
4 other projects use a version of this, 3 of them edited.
Adopted by
Most recently taken up by
- GuDaStudio/collaborating-with-codexedited, first seen Oct 9, 2026
- bkywksj/knowledge-basetheir own version, first seen Oct 8, 2026
- EarthOnlinePlayer5732/my-skillsedited, first seen Aug 11, 2026
- cnfjlhj/ai-collab-playbookedited, first seen Mar 22, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 3
- cnfjlhj/ai-collab-playbook★ 45612% same
- GuDaStudio/collaborating-with-codex★ 13226% same
- EarthOnlinePlayer5732/my-skills★ 1726% same
Wrote their own version · 1
- bkywksj/knowledge-base★ 340same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- skills/agent-memory/CLAUDE.mdCLAUDE.md
- agent-memorySkill
- auto-iterateSkill
- collaborating-with-geminiSkill
- collaborating-with-kimiSkill
- explorationSkill
- reflex-auditSkill
- reflex-fuzzSkill
- skill-catalogSkill
- skill-snapshotSkill
The file
Files
Quick Start
python3 scripts/codex_bridge.py --cd "/path/to/project" --PROMPT "Your task"Output: JSON with success, SESSION_ID, agent_messages, and optional error.
Agent Role Injection (NEW)
Inject agent roles from ~/.claude/agents/ to customize Codex behavior:
# 使用 planner agent 角色
python3 scripts/codex_bridge.py --cd "/project" --agent planner --PROMPT "Plan the auth feature"
# 使用 security-reviewer agent 角色
python3 scripts/codex_bridge.py --cd "/project" --agent security-reviewer --PROMPT "Review auth.py for vulnerabilities"
# 列出可用 agents
python3 scripts/codex_bridge.py --cd "/project" --list-agents --PROMPT ""Agent 注入优先级
--instructions- 直接传入指令字符串(最高优先级)--instructions-file- 指定指令文件路径--agent-file- 指定自定义 agent 文件路径--agent- 从~/.claude/agents/加载 agent
Parameters
usage: codex_bridge.py [-h] --PROMPT PROMPT --cd CD
[--sandbox {read-only,workspace-write,danger-full-access}]
[--SESSION_ID SESSION_ID] [--skip-git-repo-check]
[--return-all-messages] [--image IMAGE] [--model MODEL]
[--yolo] [--profile PROFILE]
[--agent AGENT] [--agent-file AGENT_FILE]
[--agent-dir AGENT_DIR] [--list-agents]
[--instructions INSTRUCTIONS] [--instructions-file FILE]
Codex Bridge with Agent Role Injection
Core Options:
--PROMPT PROMPT Instruction for the task to send to codex.
--cd CD Set the workspace root for codex.
--sandbox Sandbox policy. Defaults to `read-only`.
--SESSION_ID Resume a previous session.
--model MODEL Model to use (e.g., gpt-5.2-codex).
Agent Role Injection:
--agent AGENT Agent name from ~/.claude/agents/ (e.g., 'planner').
--agent-file FILE Custom agent file path.
--agent-dir DIR Agent directory. Defaults to ~/.claude/agents/
--list-agents List available agents and exit.
--instructions STR Direct system instructions string.
--instructions-file Path to custom instructions file.
Other Options:
--skip-git-repo-check Allow running outside a Git repository.
--return-all-messages Return all messages including reasoning.
--image IMAGE Attach image files to the prompt.
--yolo Run without approvals or sandboxing.
--profile PROFILE Configuration profile from config.toml.Multi-turn Sessions
Always capture SESSION_ID from the first response for follow-up:
# Initial task with agent role
python3 scripts/codex_bridge.py --cd "/project" --agent architect --PROMPT "Design the API"
# Continue with SESSION_ID (agent role persists in session)
python3 scripts/codex_bridge.py --cd "/project" --SESSION_ID "uuid-from-response" --PROMPT "Add error handling"Common Patterns
Planning with planner agent:
python3 scripts/codex_bridge.py --cd "/project" --agent planner --PROMPT "Plan refactoring of auth module"Security review:
python3 scripts/codex_bridge.py --cd "/project" --agent security-reviewer --PROMPT "Audit payment.py"Code review:
python3 scripts/codex_bridge.py --cd "/project" --agent code-reviewer --PROMPT "Review recent changes"Custom instructions:
python3 scripts/codex_bridge.py --cd "/project" --instructions "你是 Rust 专家,专注于性能优化" --PROMPT "Optimize this function"Debug with full trace:
python3 scripts/codex_bridge.py --cd "/project" --agent build-error-resolver --PROMPT "Fix build errors" --return-all-messagesProcess Monitoring (NEW)
使用 codex_monitor.py 监控和管理 Codex 进程:
# 列出运行中的 Codex 进程
python3 scripts/codex_monitor.py --ps
# 查看最新会话的对话内容
python3 scripts/codex_monitor.py --session latest --messages 20
# 查看指定会话(使用 SESSION_ID)
python3 scripts/codex_monitor.py --session 019c08b7-8d56-7543 --messages 50
# 实时监控最新会话(Ctrl+C 停止)
python3 scripts/codex_monitor.py --watch
# 终止指定进程
python3 scripts/codex_monitor.py --kill 12345
# 终止所有 codex exec 进程
python3 scripts/codex_monitor.py --kill all
# JSON 格式输出(便于程序处理)
python3 scripts/codex_monitor.py --ps --json
python3 scripts/codex_monitor.py --session latest --jsonMonitor 参数
| 参数 | 说明 |
--ps | 列出运行中的 Codex 进程 |
--kill <PID|all> | 终止进程 |
--session <ID|latest> | 查看会话内容 |
--watch | 实时监控最新会话 |
--messages N | 显示消息数量(默认 20) |
--json | JSON 格式输出 |
Workspace Integration (NEW)
Agent Workspace 提供任务交接与执行日志的持久化存储:
目录结构
~/agent-workspace/
├── config.json
├── projects/{project-slug}/
│ ├── plans/ # 任务计划文档
│ ├── logs/ # 执行日志 (JSONL)
│ └── artifacts/ # 中间产物
└── templates/使用方式
# 带 workspace 日志的 Codex 调用
python3 scripts/codex_bridge.py --cd "/project" \
--project swun-thesis --plan ch3-expansion --task-num 1 \
--PROMPT "Your task"
# 禁用日志
python3 scripts/codex_bridge.py --cd "/project" --workspace none --PROMPT "Quick task"Workspace Manager
# 初始化项目
python3 scripts/workspace_manager.py init --project my-project --path /path/to/project
# 查看日志
python3 scripts/workspace_manager.py logs --project swun-thesis --last 10
# 查看 plan 执行状态
python3 scripts/workspace_manager.py status --plan ch3-expansion
# 清理 30 天前的文件
python3 scripts/workspace_manager.py clean --project swun-thesis --before 30dDoes it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

