菜单
Replace with a description of the skill and when the agent should use it. Write this as a trigger condition: 'Use this skill when...'
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | template-skill |
| description | Replace with a description of the skill and when the agent should use it. Write this as a trigger condition: 'Use this skill when...' |
| license | Apache-2.0 |
| platforms | ["claude","codex","hermes"] |
| required_environment_variables | [] |
| requires_tools | [] |
| fallback_for_toolsets | [] |
| metadata | {"author":"aviskaar","version":"1.0"} |
Brief description of what this skill does.
Write your skill instructions here. Be specific about:
Describe the expected output structure.
Run a full Meridian investor intelligence report for any startup URL. Use this skill whenever the user asks to find investors, build an outreach list, write cold emails to investors, or research who to pitch. Also trigger when the user says things like "find investors for [URL]", "run Meridian on [company]", "who should I pitch", "write investor emails for [startup]", or "build my investor outreach". The skill finds 6 matched investors through real web research, verifies facts from public sources, and writes hyper-personalized cold emails grounded in actual portfolio companies, public statements, LinkedIn posts, and press releases. Never fabricate investor facts. Always search before writing.
Use this skill when you want to run an autonomous ML research loop that iteratively modifies training code, runs timed experiments, evaluates a single metric, and keeps only improvements — repeating indefinitely until manually stopped. Activate when the user wants to autonomously discover better model architectures, hyperparameters, or training strategies overnight or over a defined session, with full cross-platform support (CUDA / Apple Silicon MPS / CPU / Windows).
Use this skill when a VP of Content or content strategist needs to plan and execute an editorial strategy — producing SEO-optimized blog articles, press releases, LinkedIn posts, Medium publications, website content, and whitepapers, then distributing them across all platforms. Orchestrates seo-automation and whitepaper-engine sub-skills.
Use this skill when writing a press release for a product launch, funding round, partnership, award, executive hire, customer milestone, or any company announcement. Produces Forbes-tier, journalist-ready press releases engineered for media pickup, SEO visibility, and social virality. Covers news angle mining, headline optimization, strategic quote crafting, multimedia briefing, journalist targeting, wire distribution, and full social amplification playbook.
Use this skill when conducting a complete, end-to-end research project — from initial idea through hypothesis generation, literature review, experiment design, execution, optional replication or paper review, and final research writing. Orchestrates all available research sub-skills (hypothesis-generation, literature-synthesis, experiment-design, code-replication, research-paper-review, research-writing, ieee-paper-generator) into a single, coherent pipeline. Activate when the user wants to pursue a new research idea, replicate or extend a paper's results, or produce a publication-ready manuscript.
Use this skill when writing or drafting research paper sections using local Ollama models on GPU. Routes each section to the optimal local model — DeepSeek-R1 for methods, results, and scientific reasoning; Phi-4-reasoning for abstract, introduction, and conclusions; GLM-4.7-Flash for full drafts and cross-section synthesis. Activate when the user wants a fully local, offline paper-writing workflow without cloud APIs.
