htmlgraph

Deploy new version to PyPI using deploy-all.sh script

Diagnose orchestrator delegation enforcement gaps in the current session

<!-- Efficiency: SDK calls: 2-3, Bash calls: 0, Context: ~5% -->

Systematically capture, analyze, and track errors with HtmlGraph spike-based investigation workflow

<!-- Efficiency: SDK calls: 1, Bash calls: 0, Context: ~3% -->

<!-- Efficiency: SDK calls: 2, Bash calls: 0, Context: ~5% -->

<!-- Efficiency: SDK calls: 2, Bash calls: 0, Context: ~4% -->

<!-- Efficiency: SDK calls: 2, Bash calls: 0, Context: ~5% -->

Commit changes using the copilot-operator delegation pattern.

Display available HtmlGraph commands and usage

<!-- Efficiency: SDK calls: 1, Bash calls: 0, Context: ~3% -->

Start planning a new track with spike or create directly. Uses strategic analytics to provide project context and creates structured tracks with specs and implementation plans.

Get smart recommendations on what to work on next

Report an issue with HtmlGraph to the development team

Perform structured research before planning implementation. Uses WebSearch to gather best practices, patterns, and solutions from the web.

<!-- Efficiency: SDK calls: 1, Bash calls: 0, Context: ~3% -->

<!-- Efficiency: SDK calls: 0, Bash calls: 0, Context: ~3% -->

Create a research/planning spike

Start a new session with context continuity and status overview

<!-- Efficiency: SDK calls: 1, Bash calls: 0, Context: ~5% -->

<!-- Efficiency: SDK calls: 2, Bash calls: 0, Context: ~5% -->

Execute code generation and sandboxed tasks via OpenAI Codex CLI with automatic fallback. Use for implementation, refactoring, and structured output tasks.

Execute git and code operations via GitHub Copilot CLI with automatic fallback. Use for commits, pushes, PRs, and code generation tasks.

Systematic debugging agent. Use for error investigation, root cause analysis, and resolving issues through evidence-based methodology.

Execute research, analysis, and large-context tasks via Google Gemini CLI with automatic fallback. Use for codebase exploration, documentation research, and multi-file analysis. Free tier.

Fast, efficient code execution agent for simple tasks

Deep reasoning code execution agent for complex tasks

Research-first exploration agent. Use for understanding codebases, finding files, reading documentation, and investigating unfamiliar systems before implementing solutions.

Automated code review agent that runs roborev on recent commits and tracks findings as HtmlGraph bugs. Use after completing significant features or when asked to review recent work.

Balanced code execution agent for moderate complexity tasks

Autonomous agent for executing a single task in an isolated git worktree. Spawned by the orchestrator during parallel execution.

Quality assurance agent. Use after code changes to run tests, type checks, linting, and validate that quality gates pass.

Visual QA agent that screenshots web application pages via chrome-devtools MCP and critiques layout, readability, and correctness. Use after any UI change to validate before committing.

HtmlGraph workflow skill combining session tracking, orchestration, and parallel coordination. Activated automatically at session start. Enforces delegation patterns, manages multi-agent workflows, ensures proper activity attribution, and maintains feature awareness. Use when working with HtmlGraph projects, spawning parallel agents, or coordinating complex work.

Orchestrate parallel agent workflows using HtmlGraph's ParallelWorkflow. Activate when planning multi-agent work, using Task tool for sub-agents, or coordinating concurrent feature implementation.

Clean up git worktrees and branches after parallel execution completes. Activate when asked to clean up worktrees, remove branches, or finalize parallel work.

Code hygiene, quality gates, and pre-commit workflows. Use for linting, type checking, testing, and fixing errors.

> **DEPRECATED:** This skill is replaced by the `codex-operator` agent.

GitHub CLI (gh) operations and CopilotSpawner with full event tracking

Diagnose orchestrator delegation enforcement gaps in the current session. Use when asked "why didn't you delegate?" or to audit delegation compliance.

Execute a parallelized plan using git worktrees and multiple Claude agents. Activate when asked to run a plan, execute tasks in parallel, or start parallel development.

> **DEPRECATED:** This skill is replaced by the `gemini-operator` agent.

You are a CODER agent specialized in implementing changes efficiently. Your primary role is to modify code based on feature requirements and context provided by explorer agents.

You are an EXPLORER agent specialized in codebase discovery and analysis. Your primary role is to find, analyze, and map code without modifying it.

HtmlGraph workflow skill combining session tracking, orchestration, and parallel coordination. Activated automatically at session start. Enforces delegation patterns, manages multi-agent workflows, ensures proper activity attribution, and maintains feature awareness. Use when working with HtmlGraph projects, spawning parallel agents, or coordinating complex work.

Spawner selection, cost optimization, and HeadlessSpawner patterns for coordinating multiple AI models (Codex, Gemini, Copilot, Claude).

Cost-first delegation patterns and decision frameworks for multi-AI coordination

Monitor the status of parallel execution across git worktrees. Activate when asked for parallel status, worktree progress, or execution monitoring.

Create a parallelized development plan with research synthesis and worktree-based execution. Activate when asked to create a plan, parallelize work, or organize development tasks.

Automatically run code reviews with roborev after completing significant features and track findings as HtmlGraph bugs.

Analyze project tech stack and recommend Claude Code plugins based on detected languages, frameworks, and work patterns. Use when users ask about plugins, or proactively when you notice missing capabilities.

Use HtmlGraph analytics to make smart work prioritization decisions. Activate when recommending work, finding bottlenecks, assessing risks, or analyzing project impact.

Run visual QA on a web application — screenshots pages and reports layout, readability, and correctness issues

17 hooks across 14 events