all publishers

wuyoscar

@wuyoscar source repo

14 published skills

  1. ▌
    Aisafety Hot · wuyoscar
    Tell the user what is hottest in AI safety right now and how one of those events unfolded, from the AI Safety HOT hot board (ranked events with their reports and discussion). Use when asked "what's trending / hottest / everyone talking about in AI safety", "what happened with <event>", or for the background of a breaking story.
    0 installs
  2. ▌
    Aisafety Brief · wuyoscar
    Brief the user on what happened in AI safety in the last 24 hours or 7 days (jailbreaks, prompt injection, agent security, evals, incidents, policy), optionally filtered to their interests. Use when asked "what's new in AI safety", "this week's AI safety news", or for a morning brief on a safety topic or lab.
    0 installs
  3. ▌
    Aisafety Watch · wuyoscar
    Follow one AI lab, company or product on safety over the last 7 days — releases and system cards, research, incidents, vulnerabilities, jailbreaks, policy fights — from AI Safety HOT. Use when asked what OpenAI, Anthropic, Google DeepMind, Meta, xAI, DeepSeek or another lab did on safety lately, or whether a tool or model (Cursor, Claude Code, Copilot, an MCP server, an open-weight model) has had security problems recently.
    1 install
  4. ▌
    Aisafety Paper Scout · wuyoscar
    Find AI safety papers on a direction (jailbreaks, prompt injection, agent security, backdoors, reward hacking, interpretability, evals, AI control, privacy...) from the AI Safety HOT paper library, with one-line takeaways and BibTeX. Use when the user asks for today's new papers, papers or a reading list on an AI safety topic, or a related-work section with citations for their own paper or idea.
    0 installs
  5. ▌
    Jev · wuyoscar bundle
    Design Jev questions from a task or an existing prompt. Set up a provider, batch independent decisions, and handle custom judgments, agent checkpoints, tool routing or context checks. Read only the guide needed for this task.
    4 installs
  6. ▌
    Jev Act · wuyoscar bundle
    Choose one legal next action in a browser, desktop, game or simulation. Supply fresh observed state and available actions. The host or simulator executes and checks the result; selection does not grant permission.
    4 installs
  7. ▌
    Jev Eval · wuyoscar bundle
    Judge supplied outputs against explicit criteria, including code-change reviews and authorized safety evaluations. Use for evidence-backed review leads, rubric judgments, or batch, multi-turn and team transcript evaluation; not permission to merge or run targets.
    5 installs
  8. ▌
    Jev Triage · wuyoscar bundle
    Use for user-defined inbox, support-ticket, feedback or record classification and prioritization, especially bulk parallel judgments with sufficient per-record context. Accepts smoke_test to have the host agent write and validate a task-specific paired pilot before bulk work. Produces labels and review queues, not replies or automatic mailbox changes.
    5 installs
  9. ▌
    Jev Documents · wuyoscar bundle
    Locate, select, extract and verify evidence in documents or observed code inventories. Use for source-span extraction, passage reranking, claim checks or choosing code locations to inspect. Preserve citations and no-match outcomes; use required graph tools for exact code lookup.
    4 installs
  10. ▌
    Build Tvd Codespace 2 · wuyoscar
    Construct a new TVD (Task-Validator-Data) task and codespace from scratch, in any domain, wiring any classifier / moderation API / scorer / scientific tool into an incomplete program so a frontier model completes it and produces the data itself. Use whenever someone wants to build a new TVD task, add a template, turn a professional tool into an ISC probe, port an existing template to a new domain, or set up a TVD codespace for the agent runner. TVD is a methodology and a container, not a fixed format — every task needs its own adjustment. Covers the core idea of scale, how TVD operationalizes ISC, how to plug in any task/API/tool, how the Data file is laid out / sourced / dispensed, and the generation modes (untargeted/targeted, zero-shot/few-shot/oneshot, ICL). Enforces the one rule: the effect comes from STRUCTURE, never from keyword or prompt inducement. Keywords: TVD, ISC, build template, incomplete program, validator, anchor, trigger, classifier, moderation API, scientific tool, zero-shot, few-shot, ones
    7 installs
  11. ▌
    Arxiv2agent · wuyoscar bundle
    Digest arXiv papers into structured local folders (paper.json + per-section markdown + figures/tables/equations/listings as real files), then read and cross-check them programmatically. Use whenever the user asks to read, compare, extract from, or verify claims in arXiv papers — especially for multiple papers at once. Do NOT web-search a paper's content when its arXiv ID is known; digest it and read the structured files instead.
    3 installs
  12. ▌
    Build Tvd Codespace · wuyoscar
    Construct a new TVD (Task-Validator-Data) task and codespace from scratch, in any domain, wiring any classifier / moderation API / scorer / scientific tool into an incomplete program so a frontier model completes it and produces the data itself. Use whenever someone wants to build a new TVD task, add a template, turn a professional tool into an ISC probe, port an existing template to a new domain, or set up a TVD codespace for the agent runner. TVD is a methodology and a container, not a fixed format — every task needs its own adjustment. Covers the core idea of scale, how TVD operationalizes ISC, how to plug in any task/API/tool, how the Data file is laid out / sourced / dispensed, and the generation modes (untargeted/targeted, zero-shot/few-shot/oneshot, ICL). Enforces the one rule: the effect comes from STRUCTURE, never from keyword or prompt inducement. Keywords: TVD, ISC, build template, incomplete program, validator, anchor, trigger, classifier, moderation API, scientific tool, zero-shot, few-shot, ones
    5 installs
  13. ▌
    Get Prompt From Image · wuyoscar bundle
    Analyze user-provided reference images and reverse-engineer high-fidelity AI image-generation prompts. Use when the user asks to recreate, imitate, reverse-engineer, or extract prompts from photographs, illustrations, 3D renders, products, characters, landscapes, typography, logos, posters, or other visual references. Do not use for requests that only require OCR or an ordinary image description.
    5 installs
  14. ▌
    Gpt Image · wuyoscar bundle
    Use this skill whenever a user asks to generate, create, draw, render, or edit images with GPT Image 2 / gpt-image-2, text-to-image, reference-image editing, inpainting, posters, typography, Chinese text, UI mockups, diagrams, or gallery prompts. Analyze the user's prompt, search the bundled Reference Gallery/craft files for matching design patterns, confer on direction when useful, then call the packaged `gpt-image` CLI or bundled `scripts/generate.py`. Do not write new image-generation code unless explicitly asked to modify this repo.
    14 installs