Codex Agentic

Project management database for tracking specs, jobs, tasks, and execution. Use when Codex should run the converted pm-db workflow. Inputs: command.

Pre-planning and research for feature specifications. Use when Codex should run the converted spec-plan workflow. Inputs: feature_description.

Complete feature workflow - from planning to execution with PM-DB tracking. Use when Codex should run the converted feature-new workflow. Inputs: feature_description.

Deep research and architecture design for new product development. Use when Codex should run the converted new-product workflow. Inputs: input.

for writting unit test. Use when Codex needs this specialist perspective or review style.

Validate, critique, and iterate on generated specifications. Use when Codex should run the converted spec-review workflow.

Use this agent when you need to create comprehensive feature documentation and task breakdowns for new development work. This agent should be invoked in the following scenarios:. Use when Codex needs this specialist perspective or review style.

Use this agent when you need to design API contracts BEFORE implementation. This agent enforces contract-first API design, creates OpenAPI specifications, and defines three-tier architecture for Next.js backend APIs. Invoke in these scenarios:. Use when Codex needs this specialist perspective or review style.

Handles the visual implementation of React components. This agent is responsible for writing TSX, applying styles (CSS/SCSS/Tailwind), ensuring responsive layouts, and implementing basic user interactions (e.g., dropdowns, modals). Use this agent for any task that directly involves creating or modifying what the user sees and interacts with in the browser.. Use when Codex needs this specialist perspective or review style.

code reviews. Use when Codex needs this specialist perspective or review style.

Writes Pytest suites for unit and integration testing. Aims for 90%+ coverage. Enforces the use of Fixtures (conftest.py) over repetitive setup code.. Use when Codex needs this specialist perspective or review style.

Quick overview of Memory Bank state. Validates structure, reads all 6 files in hierarchical order, and presents formatted summary with staleness warnings.. Use when Codex should run the converted memorybank-read workflow.

Fast synchronization of activeContext.md and progress.md only. Lightweight alternative to full update for quick post-task documentation. NEW skill unique to memory-bank.. Use when Codex should run the converted memorybank-sync workflow.

Elite code review expert for Python. Focuses on Pythonic idioms, strict typing (MyPy), security (Bandit), and modern linting (Ruff).. Use when Codex needs this specialist perspective or review style.

Writes Pytest suites for unit and integration testing. Aims for 90%+ coverage. Enforces the use of Fixtures (conftest.py) over repetitive setup code.. Use when Codex needs this specialist perspective or review style.

Deep analysis of codebase for code duplication. Detects exact, structural, and pattern-level duplicates, generates comprehensive reports with refactoring suggestions and metrics.. Use when Codex should run the converted code-duplication workflow.

Resume interrupted feature work from task-list.md with PM-DB tracking. Use when Codex should run the converted feature-continue workflow. Inputs: task_list_path.

Security reviews, vulnerability scanning, OWASP compliance, and penetration testing guidance. Use when Codex needs this specialist perspective or review style.

Mode 1 - Strategic planning with human approval before execution. Use when Codex should run the converted start-phase-plan workflow. Inputs: task_list_file.

Use this agent when you need to create or improve user-facing documentation. This agent specializes in clear, accessible technical writing for API documentation, user guides, README files, tutorials, and changelogs. Invoke in these scenarios:. Use when Codex needs this specialist perspective or review style.

Read and summarize the current state of the documentation hub. Provides a quick overview of system architecture, module responsibilities, technology stack, and glossary terms.. Use when Codex should run the converted document-hub-read workflow.

Comprehensive review and update of all Memory Bank files. Reads all 6 files, proposes targeted updates focusing on activeContext.md and progress.md, and validates results.. Use when Codex should run the converted memorybank-update workflow.

Elite code review expert for Python. Focuses on Pythonic idioms, strict typing (MyPy), security (Bandit), and modern linting (Ruff).. Use when Codex needs this specialist perspective or review style.

Handles the application logic and data management for the frontend. This agent is responsible for state management (e.g., Redux, Zustand), data fetching from APIs, routing, and implementing the business logic that powers the UI components. Use this agent for tasks that are not directly about the visual presentation but about making the interface functional and efficient.. Use when Codex needs this specialist perspective or review style.

for writting unit test. Use when Codex needs this specialist perspective or review style.

Complex issue diagnosis, root cause analysis, and production incident investigation. Use when Codex needs this specialist perspective or review style.

Comprehensive review and update of the documentation hub. Analyzes recent code changes, detects drift, validates structure, and proposes specific updates to keep documentation synchronized with the codebase.. Use when Codex should run the converted document-hub-update workflow.

Initialize Memory Bank and Document Hub if not already set up. Use when Codex should run the converted documentation-start workflow. Inputs: force.

Handles the application logic and data management for the frontend. This agent is responsible for state management (e.g., Redux, Zustand), data fetching from APIs, routing, and implementing the business logic that powers the UI components. Use this agent for tasks that are not directly about the visual presentation but about making the interface functional and efficient.. Use when Codex needs this specialist perspective or review style.

Mode 2 - Structured execution with quality gates (Part 1-5) with pm-db tracking. Use when Codex should run the converted start-phase-execute workflow. Inputs: task_list_file, extra_instructions, spec_id.

Deep analysis of codebase vs documentation alignment. Detects drift, identifies undocumented code, extracts missing glossary terms, and provides actionable recommendations without making changes.. Use when Codex should run the converted document-hub-analyze workflow.

CI/CD pipelines, containerization, deployment automation, and infrastructure as code. Use when Codex needs this specialist perspective or review style.

Converted Claude specialist agent for mastra-core-developer. Use when Codex needs this specialist perspective or review style.

Bootstrap a new project's Memory Bank by creating the six core files with templates and gathering initial project information from the user.. Use when Codex should run the converted memorybank-initialize workflow.

Develops scalable FastAPI backend features. Responsible for API endpoints, Dependency Injection (Depends), Database interactions (SQLAlchemy/SQLModel), and Asyncio patterns.. Use when Codex needs this specialist perspective or review style.

Use this agent for technical debt reduction, code modernization, and architecture improvements. This agent should be invoked in the following scenarios:. Use when Codex needs this specialist perspective or review style.

Build remote control system for Claude Code using multi-agent team. Use when Codex should run the converted remote-control-builder workflow.

Bootstrap a new project's documentation hub by creating the four core documentation files (systemArchitecture.md, keyPairResponsibility.md, glossary.md, techStack.md) with initial content based on codebase analysis.. Use when Codex should run the converted document-hub-initialize workflow.

Develops scalable FastAPI backend features. Responsible for API endpoints, Dependency Injection (Depends), Database interactions (SQLAlchemy/SQLModel), and Asyncio patterns.. Use when Codex needs this specialist perspective or review style.

Automated security vulnerability scanning for Python and JavaScript/TypeScript codebases. Detects OWASP Top 10 vulnerabilities, hardcoded secrets, injection risks, and known CVEs with actionable remediation guidance.. Use when Codex should run the converted security-quality-assess workflow.

WCAG compliance, screen reader compatibility, keyboard navigation, and a11y testing. Use when Codex needs this specialist perspective or review style.

when writing backend code inside of nextjs, ie api, service intergration, database intergrations.. Use when Codex needs this specialist perspective or review style.

Parallel task execution with multi-agent teams and quality gates. Use when Codex should run the converted start-phase-execute-team workflow. Inputs: task_list_file, team_mode.

Analyze Claude Code artifacts and convert them into the appropriate OpenAI target: Codex/OpenAI skills, AGENTS.md repository instructions, or API/Agents SDK tool-based agents. Use when a repo contains CLAUDE.md files, .claude/agents, .claude/skills, or .claude/commands and Codex should first inventory the source, produce a migration plan, identify non-equivalent features, and then implement a careful conversion.

Database design, schema migrations, query optimization, and data modeling. Use when Codex needs this specialist perspective or review style.