ai_coding_project_base

Generate EXECUTION_PLAN.md and scoped AGENTS.md files for a feature. Use after /feature-technical-spec to create the task breakdown.

Generate the greenfield execution plan plus root and scoped AGENTS.md files. Use after /technical-spec to create the phased task breakdown.

Guide discovery and discussion to determine how an AI coding agent can autonomously run and verify a specific user flow end to end. Use when the user asks what would be needed to test a flow, prove an app works, create an E2E/self-test harness, make agent verification repeatable, or evaluate whether an existing test setup lets the agent verify behavior. This skill is for shaping the verification approach before implementation: read the repo's product docs and existing verification contracts, identify the real user flow and channel under test, uncover missing controllable state or external dependencies, ask focused questions, and produce a concrete agent-runnable verification plan only after the flow is understood.

Exploratory conversation before /product-spec. Asks one question at a time, runs background research for open-source/off-the-shelf alternatives, and writes DISCOVERY_NOTES.md. Use when the user has an idea but hasn't scoped it yet.

Define feature requirements (problem, users, scope, acceptance criteria) through guided Q&A and write FEATURE_SPEC.md. Use when starting a new feature.

Define technical approach (architecture, integration points, data model) for a feature through guided Q&A and write FEATURE_TECHNICAL_SPEC.md. Use after /feature-spec.

Orient to project structure and load context. Use at the start of each new session or after context reset to understand the project state.

Resume execution from wherever you left off. Detects current state and runs the appropriate next command, or reports what's blocking progress. Use at the start of any session to pick up where you left off.

Run checkpoint criteria after completing a phase. Use after /phase-start completes all tasks to verify quality gates before proceeding.

Check prerequisites before starting a phase. Use before /phase-start to verify dependencies are met and context is loaded.

Execute all tasks in a phase autonomously. Use after /phase-prep confirms prerequisites are met.

Generate PRODUCT_SPEC.md through guided Q&A. Use as the first step when starting a new greenfield project.

Generate TECHNICAL_SPEC.md through guided Q&A. Use after /product-spec to define the technical architecture.

Search past conversations with Claude Code and Codex CLI. Finds matching excerpts across all sessions with metadata (date, project, agent). Use when looking for a past discussion, decision, or piece of context from any previous session.

Analyze past Claude Code and Codex sessions to suggest new skills, workflow improvements, or changes to existing skills. Identifies repetitive patterns and unmet automation opportunities.

Check a document for AI-generated writing markers and mismatches with the user's personal tone. Two-pass analysis — generic AI detection plus voice/style comparison against the user's own writing samples from past sessions.

Initialize a new project with the AI Coding Toolkit. Use when setting up toolkit skills in a new or existing project.

Discover and sync all toolkit-using projects with the latest skills. Use when skills are modified, after the post-commit hook reminds you, or to batch-sync multiple projects.

Delegate code implementation to Codex CLI. Claude decomposes, scopes context, and verifies; Codex implements. Accepts a spec file, text description, or gathers requirements interactively.

Add a properly formatted TODO item to TODOS.md. Use when you need to capture a new task, bug, or feature request during development.

Extract automation patterns from session logs and rank by impact. Use periodically after multiple sessions to identify repetitive workflows that could become skills or hooks.

Attempt automated verification of criteria before falling back to manual. Parses criterion text for automation hints and executes appropriate tool (curl, browser, file check). Invoked by verify-task and phase-checkpoint for MANUAL and MANUAL:DEFER criteria.

Verify browser-based acceptance criteria using ExecuteAutomation Playwright MCP with multi-tool fallback chain. Invoked by verify-task and phase-checkpoint for BROWSER:* criteria.

Multi-agent code verification workflow using a main agent and sub-agent loop. Use when verifying code against requirements, acceptance criteria, or quality standards. Triggers on requests to verify, validate, or check code against specifications, checklists, or instructions.

Get a second opinion from OpenAI Codex on documents, specs, or plans. Use for cross-AI consultation on any content, not just code diffs.

Have OpenAI Codex review the current branch with documentation research. Use for second-opinion code reviews or when you want cross-AI verification.

Create a GitHub PR with verification and Codex review. Runs verify.sh and code/doc review before PR creation.

Multi-source audit of a shipped feature against its original plans, product vision, UI/UX best practices, and live browser inspection. Produces consolidated findings with prioritized recommendations.

Identify high-impact innovative additions to the current app or plan. Use when planning next features, during strategic reviews, or when the user asks for creative ideas.

Merge multiple PRs with conflict resolution and CI failure auto-patching. Analyzes optimal merge order, resolves conflicts, pulls CI logs on failure, and fixes issues. Use when multiple feature branches need merging into the base branch.

Complete OAuth login flow and store tokens for verification. Use when browser verification requires authenticated sessions.

Report completion status of the main and feature execution plans. Use to check how many tasks remain and which phases are done.

Run dependency audits, secrets detection, and static analysis to find CVEs, leaked credentials, and insecure code patterns. Use at phase checkpoints or before releases.

Verify generated specification documents for context preservation and quality issues. Automatically triggered after document generation. Checks that upstream requirements are preserved and identifies common specification problems.

Identify code duplication, cyclomatic complexity, large files, and maintainability anti-patterns. Use at phase checkpoints or on-demand to quantify technical debt.

Comprehensive UI/UX analysis and design intelligence. Point at any product to get: code analysis, product/user context assessment, industry pattern research, and prioritized improvement recommendations. Combines craft principles, data-driven design intelligence, and creative excellence. Also guides design system setup for consistent AI-assisted UI work.

Resolve Vercel preview deployment URL for the current git branch. Invoked by browser-verification when deployment.enabled is true, or directly to check deployment status. Use to check deployment status or when browser verification needs a URL.

Audits skills, commands, and prompt templates for best practice violations including length, checklists, verification steps, and progressive disclosure. Use after creating or modifying skills, during reviews, or to improve existing prompt files. Produces prioritized improvement suggestions.

Extract decisions, action items, new context, and insights from the current session into LEARNINGS.md and auto-memory. Run before ending any substantive conversation.

Auto-detect test, lint, typecheck, and build commands from package.json, Makefile, and other project config. Use during project setup or when build tooling changes. Runs silently — no prompts.