A lightweight way to divide coding work into task groups, assign each group to a coding agent, and review the results in stages. The main agent plans and checks the work while each worker handles its assigned batch.
Run a staged, evidence-first search-visibility audit of a whole live website — robots.txt and AI-crawler access, Search Console and Bing baselines, indexability, directory and review-site presence, what pages to write, and honest AI-visibility measurement. Interactive and multi-session; production files are proposed…
Generate boilerplate for new API endpoints and frontend pages following this project's conventions (flush/commit transaction pattern, ownership enforcement, domain returns). Use when: adding a new resource, creating CRUD endpoints, wiring up a new frontend page, or asking 'how do I add a new feature end-to-end.'.
Writing and editing guardrail against AI slop in German and English prose. Use whenever text is being written or revised that should not read as machine-written – LinkedIn posts, emails, newsletters, blog articles, website copy, reports, slide decks (Präsentationen, Folien, Folientitel), bios, press releases, cover…
MUST USE when user wants to research/search/look up anything on the internet — e.g. research X across the whole web / help me research X / look up X / search for X / see what people say about X / what's being discussed about X / research this topic. Also MUST USE proactively when the user: wants to watch/track/monitor…
Take a fuzzy idea (or an existing thing you want to improve) through a disciplined funnel — explore → pressure-test → converge — and produce one living CONCEPTBRIEF.md: a locked concept + phased roadmap. Acts as a blunt, honest co-founder (permission to overrule you), forces a success metric and a kill criterion…
Review the user's recent AI coding threads across Claude Code, OpenAI Codex, and Cursor to learn how THEY prefer to work — the preferences and corrections they keep repeating — and to spot places where a missing rule or script made the agent repeat work, then propose and (on approval) write those updates back into…
Writes, rewrites, reviews, and audits prompts and system prompts using Anthropic's current official guidance, and selects the right model and effort level for the task before drafting. Use when the user asks to write a prompt, improve or fix a prompt, design a system prompt, or check a prompt before running a task …
Turn a paragraph-level research brief into a research-first autonomous paper pipeline. Use for the Codex-host migration with a physically separate, plan-bound Claude Code/MiniMax M3 worker session, hash-bound evaluator evidence, authenticated lifecycle actions, typed failures, pause/resume/stop, patrol, and owned…
Two-tier project memory system. Maintains a global Claude memory snapshot of all ongoing projects AND a per-project STATUS.md file in each project folder that captures last-session state, next actions, blockers, and key locations — so context survives long absences. Operates in two modes. WRITE mode triggers when the…
Capture the current coding session's state into durable in-repo artifacts (PROGRESS.md + a self-contained HANDOVER.md) at a milestone or phase change, so a fresh session can resume without the transcript.
Use for creating or reviewing a Compose Material 3 design system whose tokens, components, previews, accessibility and performance rules fit the app context after build/tests pass.
Fetch Chinese A-share and ETF market data, including prices, valuations, capital flows, financial statements, dividends and technical indicators. Requires the ASHAREHUBAPIKEY environment variable.
Repository-agnostic Neneko CR workflow for triaging, reviewing, fixing, reporting, approving, and optionally merging GitHub pull requests across any project. It separates external-PR review from own-PR maintenance, checks current-head CI and review threads, verifies problem necessity and solution fit, requires an…
Builds a release marketing pack from README Version History: device screenshots, GenerateImage feature board and phone/web mockup posters, and a newsletter draft. Asks for brand/app icon (if not already defined) and platform (Android, iOS, or web). Use when the user says promote release, marketing posters, release…
Use when a learner wants to understand, practice, or build durable skill in a technical or analytical subject instead of outsourcing the thinking to an agent.
Opt-in only. Write, resume, or fetch/hydrate a named Cursor ↔ Claude Code handoff under .scratch/handoffs/. File first; else hydrate from the other tool's local session store (confirm required). Diff check warns if shipped ≠ tree — never blocks. Switch-only, not live sync. Do not run proactively.
A routing skill that searches a library of more than 300 skills and identifies the best match for a task. It is used when no currently installed skill fits or when someone asks to find a skill.
Reconstruct what a Steel customer organization was actually doing in its browser sessions, and whether it succeeded, from warehouse records plus raw session logs in object storage. Produces an evidence-backed report: inferred tasks, per-session outcomes, what blocked them, agent-quality problems, and what Steel itself…
★not rated 2 1mo agoA207 tokens
originalMIT
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: