bahayonghang/codex-workflow-recommender
skills/developer-tools-integrations/codex-workflow-recommender/SKILL.md
Analyze a repository and current Codex environment, then recommend Codex CLI, Codex App, AGENTS.md, skills, native subagents, plugins, MCP servers, config/hooks, and optional OMX workflow improvements without modifying files. Use when the user asks to optimize Codex workflows, improve Codex setup, recommend Codex automation, configure MCP/plugins/subagents, analyze AGENTS.md scope needs, or says "优化 Codex 流程", "Codex 工作流推荐", "给 Codex 配 MCP", "配置 Codex plugins", or "配置 Codex subagents".
$ install --global
skillsauthnpx skillsauth add bahayonghang/my-claude-code-settings codex-workflow-recommenderInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Jun 8, 2026, 2:51 AM195.1s6 files scanned
SKILL.md
- name:
- codex-workflow-recommender
- description:
- >-
- version:
- 1.0.0
- category:
- developer-tools-integrations
- argument-hint:
- [repository-or-workflow-goal]
- allowed-tools:
- Read, Glob, Grep, Bash(codex *), Bash(git *), Bash(find *), Bash(ls *), Bash(Get-ChildItem *), Bash(Get-Content *), Bash(Select-String *)
Codex Workflow Recommender
Analyze a codebase and the local Codex surface, then recommend a safe implementation order for Codex-specific workflow improvements.
This skill is read-only. It may inspect files and run read-only discovery commands, but it must not create, edit, install, remove, or reconfigure anything. End with implementation options the user can approve separately.
Output Contract
Return exactly these top-level sections unless the user asks for a narrower category:
Codebase ProfileCurrent Codex SurfaceTop Recommendations by categorySafe Implementation OrderVerification PlanWant me to implement...
Keep recommendations evidence-based. Recommend the highest-value 1-2 items per relevant category, not a catalog dump.
Discovery Workflow
1. Profile the repository
Inspect only enough to classify the project and its verification gates:
# POSIX shells
find . -maxdepth 3 \( -name package.json -o -name pyproject.toml -o -name Cargo.toml -o -name go.mod -o -name justfile -o -name Makefile \) -print
find . -maxdepth 4 \( -name AGENTS.md -o -path './.codex/*' \) -print
# PowerShell
Get-ChildItem -Recurse -Depth 3 -Include package.json,pyproject.toml,Cargo.toml,go.mod,justfile,Makefile | Select-Object -ExpandProperty FullName
Get-ChildItem -Recurse -Depth 4 -Force -Include AGENTS.md | Select-Object -ExpandProperty FullName
Capture:
| Signal | Why it matters |
|---|---|
| language/runtime manifests | hooks, test commands, subagent expertise |
| frontend/backend/data boundaries | Playwright/browser MCP, DB MCP, API docs |
| existing AGENTS.md files | root vs nested guidance recommendations |
| .codex/skills, .codex/agents | project-local Codex capabilities already present |
| ~/.codex/config.toml, ~/.codex/hooks.json when readable | user-level Codex config and hook surface |
| CI and local gates | verification plan and safe hook candidates |
| external services in dependencies or docs | MCP/plugin candidates |
| OMX files such as .omx/ or AGENTS.md OMX sections | optional repo-specific runtime workflows |
2. Confirm the local Codex command surface
Prefer the installed CLI over memory:
codex --help
codex mcp --help
codex plugin --help
codex doctor --help
When available, also inspect configured state without modifying it:
codex mcp list --json
codex plugin list
codex plugin marketplace list
If a command is missing, state that the recommendation depends on the installed Codex version.
3. Load focused references only when needed
- MCP recommendations:
references/mcp-servers.md - Skills recommendations:
references/skills-reference.md - Hooks/config recommendations:
references/hooks-patterns.md - Native subagents:
references/subagent-templates.md - Plugins:
references/plugins-reference.md
Do not load every reference by default.
Recommendation Categories
AGENTS.md guidance
Recommend root or nested AGENTS.md changes when the repository has distinct subtrees, commands, generated files, safety boundaries, or verification gates that future Codex agents must know. If the user wants direct AGENTS.md edits, use agents-md-improver instead of this read-only recommender.
Codex skills
Recommend skills for repeatable workflows, project-specific procedures, bundled scripts/templates, or domain knowledge that should be discoverable by Codex. Typical roots include:
- user/system skills:
~/.codex/skills - project-local skills:
.codex/skills - shared installs used by this repo/user environment:
~/.agents/skills
Confirm the actual roots for the current environment before giving path-specific instructions.
Native subagents
Recommend .codex/agents or ~/.codex/agents only for bounded roles with clear ownership, sandbox expectations, and verification responsibilities. Avoid suggesting a subagent when a skill, prompt, or direct instruction is simpler.
MCP servers
Recommend MCP only when external tool access materially helps: browser automation, live docs, databases, issue trackers, observability, cloud services, or filesystem boundaries. Use codex mcp add/list/get/remove syntax and avoid provider-specific commands from other CLIs.
Plugins
Recommend plugins when the user needs a bundled set of skills/tools or a reusable team distribution unit. Use codex plugin list/add/remove/marketplace syntax and verify marketplaces before recommending a plugin name.
Config and hooks
Recommend ~/.codex/config.toml or hook-related changes only when the current Codex installation and local policy support them. Treat ~/.codex/hooks.json as environment-specific; do not invent a schema when it is not present. For enforcement that must work outside Codex, recommend standard repo gates such as pre-commit, just, npm scripts, or CI.
Codex CLI runtime
Recommend direct CLI features when they match the workflow:
| Need | Codex surface |
|---|---|
| non-interactive implementation or analysis | codex exec |
| read-only diff review | codex review |
| continue or branch a previous session | codex resume / codex fork |
| diagnose install/auth/config health | codex doctor |
| sandboxed local command execution | codex sandbox |
| launch desktop app | codex app |
OMX workflows
Mention OMX only when the repository or user environment shows OMX is installed or requested. Present it as an optional enhancement, not a universal Codex capability.
Report Template
## Codex Workflow Recommendations
### Codebase Profile
- **Type**: ...
- **Primary gates**: ...
- **Risk boundaries**: ...
### Current Codex Surface
- **AGENTS.md**: root/nested/global status
- **Skills**: discovered roots and notable project-local skills
- **Subagents**: discovered `.codex/agents` or user agents
- **MCP**: configured or absent
- **Plugins**: configured marketplaces/plugins or absent
- **Config/hooks**: current known surface and unsupported gaps
- **OMX**: present/absent/optional
### Top Recommendations by category
#### AGENTS.md
1. **...**
- Evidence: ...
- Why now: ...
- Suggested scope: ...
#### Skills
...
#### Native subagents
...
#### MCP servers
...
#### Plugins
...
#### Config/hooks
...
#### Codex CLI runtime
...
### Safe Implementation Order
1. Read-only inventory and baseline verification
2. AGENTS.md or docs guidance changes
3. Local/project skill or subagent additions
4. MCP/plugin/config changes behind explicit approval
5. Verification and rollback notes
### Verification Plan
- `codex doctor` for installation/config health
- repository-specific lint/type/test/build gates
- skill or agent validators when new packages are added
- `codex mcp list --json` / `codex plugin list` after configuration changes
### Want me to implement...
I can implement the approved local-file changes next. I will not install plugins, add MCP servers, or change user-level config without explicit approval.
Safety Rules
- Do not modify files or run install/remove/config mutation commands in this skill.
- Do not claim a Codex feature exists solely because another CLI has it.
- Do not recommend secrets in config files; prefer environment variables and bearer-token env var options where supported.
- Do not place OMX recommendations in the critical path unless the current environment actually uses OMX.
- Keep implementation advice reversible and ordered from lowest-risk to highest-risk.
Related Skills
bahayonghang/literature-mentor
tools
VerifiedTrustedCommunity
文献深度解读助手，像研究生导师一样交互式解读 Zotero 库中的学术论文，面向计算机科学、深度学习、自动化等方向（个人向）。当用户提供文献题目、DOI、PDF 或要求解读某篇论文时触发，通过 Zotero MCP 优先获取全文，并根据用户意图自动选择快速筛选、导师深读或研究复盘模式。完整深读时先完成叙事类型判断、阅读前预检、novelty 校准和作者思考路径重建，再整体概览，并基于图例、正文和表格逐图详细解读（Zotero MCP 无法提取 PDF 图片，解读基于文字信息，必要时提醒上传图片）。适用于：(1)快速判断文献是否值得深读 (2)深入理解某篇论文 (3)学习文章中的方法和技术 (4)批判性分析研究设计 (5)寻找研究灵感。需要多篇论文综合、对比或找研究空白，或 arXiv/DOI 批量规范化时，改用 paper-workbench。
16SKILL.mdUpdated Jun 21, 2026
bahayonghang/agent-skill-review
development
VerifiedTrustedCommunity
Review Codex, Claude, OpenAI, or other agent skill directories as reusable capability packages. Use when asked to audit, review, improve, score, rewrite, debrand, package, or document a SKILL.md, skill package, marketplace skill, or agent skill directory, especially when the user wants a comprehensive findings-first report with concrete patch recommendations and validation steps.
16SKILL.mdUpdated Jun 15, 2026
bahayonghang/goal-meta-skill
development
VerifiedTrustedCommunity
Turn vague or complex Codex tasks into strong `/goal` commands with outcome, verification, constraints, boundaries, iteration policy, completion evidence, and pause/block conditions. Use when the user asks for Codex goal instructions, Goal 指令, 目标指令, `/goal` prompts, 中文 Goal 模板, plan-to-goal interviews, success criteria, verification commands, or bounded agent work definitions.
16SKILL.mdUpdated Jun 13, 2026
bahayonghang/ast-grep
tools
VerifiedTrustedCommunity
Write, debug, and validate ast-grep structural code search rules. Use this skill when the user needs syntax-aware code search, AST pattern matching, structural refactor discovery, language-construct queries, or searches that plain text tools like rg can miss, such as finding functions with particular descendants, calls inside specific contexts, missing error handling, React hook shapes, decorators, or other Tree-sitter-backed code structures.
16SKILL.mdUpdated Jun 9, 2026