costiash

Queries GPT-5.2 for high-reasoning code analysis, root-cause bug fixing, and complex coding questions. Provides P0-P3 prioritized analysis reports, architecture audits, and security reviews with configurable reasoning effort (none/low/medium/high/xhigh). 400K context, 128K output.

Queries Gemini 3 Flash for high-speed code analysis, generation, and complex coding questions. Provides P0-P3 prioritized analysis reports, architecture audits, and code generation with configurable thinking levels (minimal/low/medium/high). 1M context, 64K output. Pro-level intelligence at Flash pricing.

Guides users through video transcription workflow from input to output. Transcribes local video files and YouTube URLs using gpt-4o-transcribe. Use when users want to transcribe videos, audio files, YouTube content, or need help with media-to-text conversion.

Identifies and merges duplicate entities in Knowledge Graph projects. Use after extraction to consolidate duplicate entities, or when users ask about potential duplicates. Supports automatic merging for high-confidence matches and user confirmation for medium-confidence candidates.

Creates and bootstraps Knowledge Graph projects from video transcripts. Extracts entities (people, organizations, concepts) and relationships into searchable graphs. Use when users want to build a knowledge graph, extract entities, analyze connections between people/organizations, or select Option 5 after transcription.

Helps users discover patterns and insights in their Knowledge Graphs. Use when users ask about important entities, connections, patterns, or want to understand what they can do with their graph. Triggers proactively after extraction to show what's possible.

Guides users through saving generated content (summaries, notes, key points) to professionally formatted and themed files. Use when users want to save, export, or persist content generated during the session, choose Option 4 after transcription, apply styling/themes to saved content, or explicitly ask to save summaries, notes, or derived content.

Multi-model AI council for actionable project advice. Leverages Gemini 3 Flash and GPT-5.2 skills in parallel, then synthesizes through an Opus Judge for stage-appropriate, non-overkill recommendations. Use when seeking architectural guidance, code review synthesis, or implementation planning.

Generate a self-contained HTML changelog report showing recent documentation changes with interactive course generation buttons. Use this skill when the user asks for a docs changelog, documentation report, what changed recently, recent doc updates as a report, or runs `/docs --report`. Discovers changes via git history, categorizes them, summarizes what's new, and produces a stunning Obsidian & Amber themed HTML report where each entry has a "Create Course" button that copies the course command to clipboard.

Generate an interactive, self-contained HTML course on any Claude documentation topic. Use this skill when the user wants to create an interactive course, tutorial, or deep-dive walkthrough about a Claude feature, API concept, SDK pattern, or prompt engineering technique. Triggers on: "create a course about", "interactive tutorial for", "teach me about X interactively", "make a course on", "I'd like a course", "/docs --course", "/docs course", or when the user accepts the post-docs course prompt. Produces a stunning single-page HTML file with scroll-based navigation, animated visualizations, quizzes, and code translations drawn from official Claude documentation.

Search and read locally-stored Claude documentation covering Claude Code CLI, Claude API (Messages, tool use, vision, streaming, batch), Agent SDK (Python and TypeScript), prompt engineering, and all Anthropic platform docs. Use this skill whenever the user asks about Claude Code features (hooks, MCP servers, skills, plugins, settings, permissions, keybindings, sub-agents), the Anthropic API or any of its SDKs (Python, TypeScript, Go, Java), the Agent SDK (sessions, hooks, custom tools, MCP), model capabilities (context windows, extended thinking, pricing, rate limits, vision), prompt engineering best practices, or troubleshooting any Claude-related error. This skill provides instant access to official documentation files without web searches — always prefer it over web lookups for Claude and Anthropic topics.

Check the health and freshness of locally-stored Claude documentation. Use this skill when the user asks about documentation health, broken links, stale docs, freshness checks, or wants to validate that their local mirror is up-to-date and all URLs are reachable. Triggers on: "are my docs current", "check doc health", "validate documentation", "broken links", "stale docs".