Opus-based critical thinking agent for challenging assumptions, narrowing decision options, and deeply reviewing plans or results. Use BEFORE major decisions (architecture, design, strategy), when comparing multiple options, or when you need a devil's advocate perspective on a plan or deliverable. Returns structured…
Memory search and recall specialist. Use when you need to find past decisions, conversations, project history, learned workflows, or any stored knowledge from the agent's memory system. Returns structured summaries.
Analyzes a Claude Code session JSONL file and extracts structured insights for memory consolidation. Used by the dreaming trigger to process each session independently.
Use this agent when reviewing code that involves error handling, catch blocks, fallback logic, or retry patterns. Specifically use when PRs contain try-catch blocks, error callbacks, fallback chains, or any code that could silently suppress errors. Complements security-code-reviewer by focusing on error handling…
Use this subagent for SMALL, FOCUSED video editing tasks on existing footage — trimming, concatenating, removing silence/fillers, adding subtitles, simple color correction, audio normalization, watermarks, format conversion. The subagent knows the audio-first AI-driven editing workflow (transcribe → EDL → FFmpeg…
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a…
Choose or review visual hierarchy, typography, color and layout for interfaces, documents, presentations and graphics. Use existing brand and project conventions when extending a design.
Extract text or OCR from documents and images for reading or summarization. Use format-specific skills when formulas, tracked changes, speaker notes, or editing must be preserved.
Create or edit Word documents, work with templates, comments and tracked changes, or inspect Word-specific structure. For plain text extraction or summarization, use document-parse.
Implement web pages and components with responsive layouts, interaction states and accessible markup. Use design for shared visual decisions and preserve existing project styles.
Use this skill for any PDF authoring or post-processing task that doesn't require READING the content. CREATE / GENERATE / PRODUCE PDFs from scratch — reports, invoices, business letters, memos, status updates, market analyses, proposals (Typst engine, charts via Cetz, four bundled templates: report, invoice, letter…
Create or edit PowerPoint presentations, inspect slide layouts and speaker notes, or combine decks. For plain text extraction or summarization, use document-parse.
Task management with the task CLI. Use when planning multi-step work, tracking goals with done-conditions, managing dependencies, or gating session exit on completion. Triggers on 'create task', 'track progress', 'plan work', 'open goals', 'what tasks are open', or any work that benefits from structured decomposition.
How to create and manage triggers via the CLI. Use for durable cron schedules, incoming webhooks, manual jobs, or messaging integration setup. Use reminders for one-shot follow-ups.
Cut AI tells from any writing. Must always apply to text a human reads, including chat and Signal replies, emails, documents, reports, and commit messages.
Edit existing footage into a finished video: choose takes, cut speech or silence, grade, add overlays and subtitles. Uses timestamped transcripts and bundled FFmpeg helpers. Use video for Remotion animation authoring.
Create or edit Remotion videos, animations and compositions, render them, or build Remotion players. For cutting existing footage with FFmpeg, use video-edit.
Writing documents an agent reads. Use when creating or editing a skill, a trigger prompt, a subagent task prompt, a memory file, or CLAUDE.md / AGENTS.md.
Create, edit, or analyze spreadsheets and tabular files while preserving cells, formulas and data types. For a plain text summary of a document, use document-parse.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: