Ferramentas

Guides the agent through writing or improving a skill: clarify intent, draft SKILL.md, run scored evals with and without the skill, then tune the description so it triggers accurately. Official Anthropic skill.

166 research skills maintained by K-Dense (formerly Claude Scientific Skills), spanning bioinformatics, cheminformatics, clinical research, materials and physics, including unified access to 78 public databases such as PubChem and UniProt. Works with Claude Code, Codex and Cursor.

Reference for the Claude API and Anthropic SDK: model selection and pricing, streaming, tool use, prompt caching, Managed Agents and more, plus workflows for model migration and prompt audits. Official Anthropic skill.

Guides agents in building quality MCP servers: research the protocol and target API, design tools, implement with Python (FastMCP) or the TypeScript SDK, and write evaluation questions to test the result. Official Anthropic skill.

Completes and populates three-statement financial model templates, linking the income statement, balance sheet and cash flow with formulas and checking that the balance sheet balances and cash ties out, with no hardcoded values. Part of Anthropic's official financial-analysis plugin.

Builds M&A accretion/dilution models: sources and uses, pro forma EPS and goodwill amortization impact, with sensitivities on synergies, premium and cash/stock mix, in an Excel workbook. You supply share prices, EPS and deal terms. From Anthropic's investment-banking plugin; needs expert review.

Anthropic's official financial services skills for investment banking, equity research, private equity, fund admin and wealth management: comps, DCF, LBO, three-statement models, CIMs, earnings reviews, IC memos, GL reconciliation and more. Outputs need expert review.

Official Anthropic earnings skill: for a covered company, reads the quarterly report and call, compares results with consensus for a beat or miss, breaks down key metrics, updates estimates and thesis, and writes an 8-12 page update with charts. Needs analyst review; not investment advice.

Official Anthropic private equity skill: turns company, industry, financials, deal terms, diligence findings and returns into a standard IC memo with executive summary, key risks and mitigants, base, upside and downside returns and a clear recommendation, as a Word doc by default.

Official Anthropic initiating coverage skill: works in five separate steps (company research, financial model, valuation, charts, final report) to produce an Excel model, 25-35 charts and a 30-50 page Word report styled after bulge-bracket sell-side research. Needs analyst review.

Official Anthropic finance skill that checks Excel for formula errors like #REF!, numbers hardcoded in formulas, sum ranges that miss rows and mixed units; on full models it also checks the balance sheet balances and cash ties. Lists issues and fixes and asks before changing anything.

Builds a DCF valuation in Excel: pulls figures from SEC filings and research, projects free cash flow, computes WACC and terminal value, and adds sensitivity tables. Derived values are formulas, inputs cite sources, and each step is confirmed with you. From Anthropic's financial-analysis plugin.

Official Anthropic LBO modeling skill: builds sources and uses, operating projections, debt schedule and IRR/MOIC returns in an Excel template, with every calculation as a formula and each section checked with you. Works in the Excel add-in or generates files with local Python; needs expert review.

Official Anthropic comparable company analysis skill: selects listed peers, compiles operating metrics and multiples like EV/EBITDA and P/E with percentiles, and outputs a comps table in Excel for valuation, IPO or financing pricing. Prefers paid sources like FactSet but can use SEC filings.

Official Anthropic fund admin skill: matches general ledger and subledger exports for a date or period line by line, lists mismatched amounts, quantities, posting dates and one-sided entries, suggests causes such as timing, FX or mapping, and writes a break report. Needs accountant review.

When a user asks whether a skill exists for some task, searches the skills.sh ecosystem with the skills CLI, then recommends and installs a fit. Maintained by Vercel Labs and shipped in the skills CLI repo.

Helps finance teams with ratio analysis, DCF valuation, budget variance analysis and rolling forecasts, using bundled Python scripts so the numbers can be recomputed. Needs Python locally and financial data prepared as JSON; have a professional review the results.

Google's official skill library, focused on Google Cloud: 100+ skills for gcloud, Cloud Run, GKE, BigQuery, IAM, monitoring and logging, and the Well-Architected Framework, plus Ads and Analytics.

Context and harness engineering skills: context fundamentals, degradation diagnosis and compression, multi-agent architectures, memory systems, tool design, LLM-as-judge evaluation and self-improvement loops. Focused on principles and design patterns, not tied to any agent platform.

Lets Claude Code push ML research forward unattended: literature review, topic selection, experiment deployment, cross-model review loops and paper writing through rebuttal, in 80+ skills. Codex acts as the independent reviewer by default; other model pairings work too.

Value investing research skills that turn the methods of Buffett, Munger, Duan Yongping and Li Lu into 20+ workflows: deep stock research, filing reviews, sector screens, portfolio reviews and thesis tracking, with clear conclusions and cross-checked data. Token-heavy. Not investment advice.

Writes precise prompts for a specific AI tool: identifies the target tool, extracts your intent, asks at most three questions if details are missing, then outputs a lean, paste-ready prompt. Covers dozens of tools including ChatGPT, Claude Code, Cursor, Midjourney and Sora.

Nearly 100 skills spanning the AI research workflow: ideation and literature review, fine-tuning and post-training, distributed training, inference serving, evaluation, interpretability and paper writing, tied together by an autoresearch orchestration skill.

Teaches agents the Hugging Face Hub CLI hf: search and download models, manage datasets, Spaces, Buckets and repos, and run Jobs and inference endpoints. The skill content is generated from your local CLI, so install hf first. Maintained by Hugging Face.