anthropics/financial-services-plugins
27 skills21.4k installs823k starsGitHub
Install
npx skills add https://github.com/anthropics/financial-services-pluginsSkills in this repo
1Earnings AnalysisEarnings Analysis is a financial-services agent skill oriented toward writing crisp post-earnings updates for listed companies. It encodes how research-style headlines should read—company, period, thesis in one line—and how body copy should quantify beats and misses, attribute results to drivers, refresh estimates when moves are material, and tie commentary to management calls and peer prints. Solo builders building fintech dashboards, investor newsletters, or internal research copilots install it when they need repeatable quality gates on earnings copy rather than generic summarization. The embedded examples contrast strong versus weak headlines and list operational tips such as publishing within 24–48 hours and updating price targets when estimates shift materially. It is less about building models from scratch and more about editorial and analytical consistency on quarterly results.1.5kinstalls2Equity ResearchEquity Research is an agent skill for solo builders and analysts who need institutional-style snapshots without manually juggling terminals. It instructs the agent to act as an equity research analyst: call the documented MCP tools for IBES consensus, reported fundamentals, historical prices, and macro backdrop, then weave outputs into a structured investment narrative. The emphasis is thesis-first—each table should support whether market expectations look too optimistic or too pessimistic—not raw data dumps. Use it when researching a ticker, comparing estimates to actuals, stress-testing valuation, or drafting a memo for a fintech side project or content product. It fits builders wiring financial MCP plugins into Claude or similar agents who want repeatable sectioning and table discipline instead of one-off chat answers.1.4kinstalls3Macro Rates MonitorMacro-rates-monitor is an agent skill from the financial-services plugin set that turns MCP-sourced macro and fixed-income data into strategist-grade dashboards and commentary. It is aimed at solo builders, analysts, or indie fintech builders who already expose qa_macroeconomic, interest_rate_curve, inflation_curve, and swap-rate tools to their agent and need disciplined synthesis—not raw series dumps. The skill encodes core principles: place the economy in the cycle using GDP, employment, and PMI; interpret central bank policy via policy rates and curve shape; read bond-market signals from slope and real rates; and judge whether financial conditions are tightening or easing. You invoke it when monitoring macro conditions, analyzing yield curve shape, splitting real versus nominal rates, mapping policy rate expectations, or evaluating financial conditions. The agent’s job is routing and narrative coherence while tools supply numbers.1kinstalls4Competitive AnalysisCompetitive Analysis is a financial-services agent skill for producing rigorous competitive landscape decks instead of vague bullet lists. It targets solo founders, indie operators, and small teams who need investor-ready or internal strategy slides covering market positioning, competitor deep-dives, comparative metrics, and synthesis. The workflow is deliberately gated: Phase 1 uses structured questions to distinguish a five-slide market map from a twenty-slide peer benchmark before any research spend. Phase 2 builds into an open PowerPoint add-in session or delivers a .pptx in chat. Use it when someone asks who competes with a company, how peers benchmark, or for a strategic review packaged as slides. The skill emphasizes not guessing scope because competitive analysis shapes differ widely. Intermediate complexity reflects finance literacy plus deck mechanics rather than code.973installs5Ppt Template CreatorPPT Template Creator is a meta agent skill for solo builders who already have a corporate or course PowerPoint template and want to turn it into a distributable Claude/Cursor skill. It is only for creating reusable skill packages: you supply a .pptx or .potx, the agent analyzes slide layouts and placeholders (with python-pptx), initializes structure through skill-creator, copies the file into assets, writes a complete SKILL.md from the provided template pattern, validates with a sample presentation, and packages a .skill file. It deliberately does not render finished pitch decks—that is the job of a dedicated pptx skill generated or used afterward. Financial-services and indie founders use it when brand consistency matters and they plan to generate many decks from the same master. Pair it with skill-creator for conventions; refer to skill-creator for general authoring best practices beyond PPT-specific steps.950installs6Dcf Modeldcf-model equips a coding agent to construct and audit discounted cash flow spreadsheets used in financial services workflows. The embedded validator loads the same workbook twice—with and without formula evaluation—to catch broken references, structural gaps, and common DCF modeling mistakes before stakeholders trust the output. Solo founders validating a B2B fintech idea, indie analysts packaging diligence deliverables, or agent builders extending Anthropic financial plugins can invoke it when an .xlsx model needs a systematic pass rather than eyeball review. Outputs consolidate into JSON-friendly status with separated errors and warnings so you can gate sharing or iteration. It assumes Python with openpyxl installed locally and a concrete Excel path; it is not a substitute for licensed market data feeds unless you wire requests-based sources yourself. Use when pricing a venture, updating terminal value logic, or preflight checks before investor review.914installs7Pitch Deckpitch-deck is a financial-services agent skill focused on calculation verification for investor-facing decks. Solo founders and small teams often paste third-party market reports or advisor models into slides without reconciling CAGR paths, implied enterprise value from multiples, or share math. This skill supplies reference formulas—compound growth projection, EV over revenue or EBITDA, and market share identities—plus rounding rules, a verification checklist, and explicit red flags when claimed figures do not close. Use it when you already have populated or semi-populated deck data and need the agent to sanity-check arithmetic before a fundraise conversation, partnership pitch, or internal board readout. It does not replace legal or audit counsel; it tightens the quantitative layer so narrative claims match defensible inputs. Best paired with your own source citations and template workflow during Validate and again before Launch materials ship.849installs8Option Vol AnalysisOption Vol Analysis is an Anthropic financial-services agent skill for derivatives volatility work. Solo quant builders and indie fintech agents install it when they need disciplined routing: start from the implied vol surface the market embeds, drill into specific options for Greeks via pricing tools, then benchmark against realized vol from historical series so the vol premium tells you whether protection is rich or cheap. It targets MCP-connected workflows for equities, indices, and FX—not generic chat guesses. Use during Operate when you iterate hedges, roll expiries, or sanity-check agent-generated trade ideas before capital goes live. Prerequisites include working MCP financial data connectors and comfort reading surface shape (skew, term structure). It does not replace compliance review, execution algos, or position limits, but it standardizes how agents chain surface, price, and history steps for repeatable vol assessments.818installs9Comps Analysiscomps-analysis is a finance agent skill for building comparable company tables the way institutional teams expect: peer operating metrics, trading and transaction multiples, and statistical context exported for Excel workflows. Solo founders, indie fund analysts, and small investment groups use it when they must defend a valuation band, benchmark growth versus public peers, or frame an IPO or round—not when guessing from blog posts. The skill enforces a strict data-source order: use connected financial MCPs when available and avoid web search as a primary source, preserving auditability for committee-ready work. It shines on liquid public peer sets; the skill itself warns away from private comps with no public analogs, conglomerates, distressed credits, and pre-revenue startups. Output is meant to support presentations and pricing decisions during validation, not to replace legal or compliance review.806installs10Fixed Income PortfolioFixed-income-portfolio is an advanced agent skill for financial-services workflows that turns MCP bond tooling into structured portfolio reviews. It instructs the agent to act as a fixed income analyst: batch price identifiers for clean and dirty prices, yields, duration, convexity, DV01, and spreads; pull issuer, rating, sector, and call data for composition; build cashflow waterfalls for reinvestment risk; and run rate scenarios for stress testing. The skill emphasizes letting tools compute bond-level analytics while the agent aggregates to portfolio metrics and narrates risk relative to benchmarks. It targets builders and operators shipping Claude-based fintech copilots rather than casual indie app makers, and assumes MCP servers for pricing and Yieldbook-style reference and cashflow APIs are already connected.733installs11Clean Data XlsClean Data XLS is a procedural agent skill for solo builders and operators who live in spreadsheets—especially finance-adjacent exports that look fine visually but break pivots, joins, and charts. It scopes either a named range or the sheet’s used range, profiles each column’s dominant type, then applies a documented issue matrix covering whitespace, casing drift, numbers stored as text, ambiguous dates, and duplicate rows. The environment split matters: inside Excel as an add-in it reads and writes through Office JS, including helper-column formulas like TRIM when in-place edits are risky; on detached files it uses openpyxl for the same hygiene goals. You invoke it when a user says the data is messy or before analysis, not as a substitute for a full ETL pipeline in production warehouses. It pairs naturally with downstream charting or modeling steps once types are consistent and duplicates are resolved.732installs12Earnings Preview SingleEarnings Preview Single is a catalog entry for an Anthropic financial-services plugin skill aimed at builders who ship finance-adjacent agents or internal tools for analysts and operators. Public SKILL text in the ingest is license-forward only; placement assumes the standard pattern for sibling earnings and research plugins—one issuer preview document or brief rather than a full sector screen. Solo builders install it when they need repeatable, agent-invoked earnings prep for a single ticker or company event. Pair it with your data feeds and compliance review; do not treat model output as investment advice. Confidence is moderate until full SKILL.md workflow steps are present in the bundle you install.731installs13Initiating CoverageInitiating-coverage is a financial-services agent skill that enforces institution-style completeness on equity initiation reports before you hand work to stakeholders or clients. Solo builders and indie analysts who use Claude or similar agents to draft coverage can treat it as a final QA layer: it walks through deliverable files, page and word counts, chart formats, and model attachments so outputs do not look like thin AI drafts. The skill fits when you are producing initiation-style research artifacts—not when you only need a quick memo. It pairs naturally with agent workflows that generate DOCX and spreadsheets, and it makes expectations explicit (actual image charts, sensitivity grids, product and geography revenue views) so downstream readers trust the package. Confidence is high for checklist fidelity; you still own data sourcing and compliance with your jurisdiction’s research rules.730installs143 Statement Model3 Statement Model is a finance agent skill from Anthropics’ financial-services plugins that encodes how to structure and format a classic integrated financial model—profit and loss, balance sheet, and cash flow—with spreadsheet conventions investors and operators expect. Solo and indie builders use it when they need forecasts that tie together revenue assumptions, working capital, debt, and equity without silent formula drift or ambiguous presentation. The ingested reference emphasizes presentation and integrity mechanics: color-coded inputs versus formulas, parentheses for negatives, header and units rows, bold aggregates on named total lines, and visual breaks between historical actuals and forward projections. It is not a substitute for your domain assumptions; it is the scaffolding so your agent does not ship a model that fails balance checks or hides hardcodes. In Prism’s journey, it anchors Validate when you are pricing and modeling runway, with follow-on value in Operate when refreshing actuals and in Idea when framing market-size back-of-envelope links—always as spreadsheet craftsmanship layered on your business logic.724installs15Audit XlsAudit XLS is an agent skill that systematically reviews spreadsheets for formula errors, bad practices, and—at model scope—financial-statement integrity. Solo founders, analysts, and indie operators use it when a model feels wrong, won’t balance, or must pass a last-mile check before a client deck or IC readout. The workflow starts by locking audit depth (selection vs sheet vs full model), then applies a consistent formula checklist everywhere and deepens into BS balance, cash reconciliation, and roll-forward logic when the whole workbook is in scope. It reduces the risk of shipping silent hardcodes, broken references, or structural accounting mistakes that spreadsheet tools alone rarely explain in plain language.709installs16Bond Futures Basisbond-futures-basis is an agent skill for fixed-income builders and quant-minded solo developers who connect Anthropic financial-services MCP tools to analyze government bond futures basis trades. It applies when you need fair futures pricing, cheapest-to-deliver selection, implied repo versus market repo, and a readable narrative on whether futures look rich or cheap—not when you want the model to invent prices without tool output. The skill encodes expert routing: start from futures pricing and delivery mechanics, price the CTD cash bond separately, then synthesize basis metrics and curve context. That discipline reduces inconsistent hand-waved basis commentary in agent chats. Typical use is prototyping fintech research agents, internal trading assistants, or education flows around Treasury futures. Advanced familiarity with repo, conversion factors, and delivery options is assumed; tools compute, the agent interprets and presents.702installs17Lbo ModelLBO Model is an agent skill for private equity and corporate finance workflows where Claude completes leveraged buyout spreadsheets instead of inventing layout from a blank workbook. Solo builders and small deal teams use it when they already have—or can attach—a firm template and need assumptions wired into Sources & Uses, cash-flow projections, debt schedules, and IRR or MOIC returns with consistent formatting. The skill opens by checking for an attached Excel file and treats that structure as authoritative; if none is supplied it offers the bundled LBO_Model.xlsx example or asks for the user’s standard model. It emphasizes formula integrity and presentation suitable for investment committee packets rather than quick back-of-envelope math. Environment detection matters because Excel add-in Office JS automation differs from Python-based file manipulation. This is specialized finance tooling—not a general spreadsheet assistant—best when transaction inputs are known and the deliverable is a defensible LBO model.688installs18Bond Relative ValueBond Relative Value is an agent skill for fixed-income analysts and indie fintech builders who want repeatable richness/cheapness workflows inside Claude Code or similar agents. It assumes MCP tools such as bond_price, interest_rate_curve, and credit_curve are available, and it teaches the agent to combine pricing, government and swap curves, and issuer credit curves into a coherent relative value narrative. The workflow emphasizes spread decomposition—splitting compensation into rate, credit, and residual components—because the residual is where mispricing versus comparables usually shows up. You use it when comparing ISIN, RIC, or CUSIP-identified bonds, assessing value versus curves, or running rate shock scenarios before sizing or hedging. Complexity is advanced: you need curve conventions, spread definitions, and comfort delegating arithmetic to tools while owning interpretation. It is not a general coding skill; it is procedural knowledge for credit and rates desks, wealth-tech prototypes, or solo builders shipping bond analytics copilots.679installs19Fx Carry TradeFX Carry Trade Analysis is an agent skill for quantitative FX strategists and indie builders wiring Anthropic financial MCP tools into repeatable carry reviews. It starts from current spot, pulls forward points and outright rates at target tenors, walks the forward curve to compare roll and tenor choice, and combines interest-rate differential logic with volatility surface context so annualized carry can be compared to ATM implied vol—the carry-to-vol ratio the doc treats as the headline risk-adjusted metric. The skill emphasizes that carry earns the rate differential while spot risk dominates P&L, and that carry structures are naturally short volatility, so vol regime shifts matter as much as rate spreads. Install it when you are analyzing carry trades, comparing FX forward curves, assessing carry-to-vol ratios, or screening currency-pair opportunities rather than ad-hoc spreadsheet copy-paste. It expects MCP-backed pricing tools to return mid/bid/ask and curve data; you interpret and synthesize. Not a general macro narrative generator—it is a structured routing layer from market data into carry recommendations.670installs20Datapack Builderdatapack-builder is a financial-services agent skill that walks solo builders and small deal teams through extracting, normalizing, and formatting financial data into polished Excel workbooks suitable for investment committee review. It expects messy inputs—confidential information memos, offering memorandums, SEC filings, web research, or MCP-connected sources—and enforces professional structure, documented assumptions, and formula-driven models instead of pasted numbers. The skill stresses traceability for every figure, internal consistency across statements, and balance sheet identity checks so a single bad subtotal does not torpedo a diligence read. It is not for quick arithmetic on finished models; it is for producing repeatable data packs across deals or portfolio companies. Throughout the workflow you should delegate spreadsheet mechanics to the xlsx skill while this skill governs financial rigor and IC-ready presentation.666installs21Tear SheetTear Sheet is an agent skill from the Anthropic financial-services plugins line, intended to help produce the kind of one-page investment or instrument summaries professionals share before deeper diligence. Prism ingested primarily license text for this entry, so treat capability details as plugin-conventional: condensed fundamentals, narrative hooks, and tables suited to wealth, asset management, or fintech copilots—not a replacement for compliance-reviewed offering documents. Solo builders shipping finance-adjacent SaaS can invoke it while validating positioning, drafting demo data rooms, or scripting copilot features that must output advisor-grade PDFs or HTML snippets. Always layer your own data sources, disclaimers, and jurisdiction rules on top of model-generated sheets. Re-open the full SKILL.md in the plugin repo before production prompts that touch MNPI, performance claims, or client-specific advice.658installs22Swap Curve StrategySwap Curve Strategy is an agent skill for solo builders and small quant teams who wire financial MCP tools into Claude-class agents to behave like a rates strategist. The workflow insists on sound sequencing: price swaps at multiple tenors to form the swap curve, layer government and inflation curves, compute swap spreads and real-rate components, then interpret shape via slope and butterfly metrics with historical context. It is explicitly for analyzing swap curves, computing swap spreads, decomposing real rates, and identifying steepener, flattener, and butterfly opportunities—including cross-currency comparisons when data allows. You stay in analysis and recommendation mode while tools handle pricing math. This is niche relative to typical indie SaaS journeys but essential if you are validating a fintech agent, internal treasury assistant, or research copilot. Invoke when user questions mention swap curves, swap spreads, or curve trades rather than equity or credit single-name stories.654installs23Fsi Strip ProfileFSI Strip Profile is an agent skill that walks you through creating professional investment banking strip profiles for pitch books, deal materials, and client presentations. It is built for solo analysts, indie consultants, and small teams who must turn filings and market data into one to four information-dense slides with quadrant layouts, charts, and tables—without skipping the IB discipline of confirming slide count and focus up front. The workflow starts by asking whether you need a single slide or a multi-slide set and which topics to emphasize, then directs research across company filings, investor presentations, consensus estimates, and recent news before any build step. You normalize currency and scale, assemble revenue, EBITDA, margins, EPS, FCF, valuation, ownership, and segment breakdowns, and print a numbered outline with actual figures for approval. Style choices such as fonts and hex colors are decided explicitly so output matches firm or client branding expectations.653installs24Funding DigestFunding-digest is an agent skill from Anthropic’s financial-services plugins family, intended to help builders stay oriented on funding and capital-flow context while exploring financial or fintech ideas. Prism’s ingested readme is license text only, so treat the skill as a research-oriented digest helper you invoke when you want your agent to condense funding-relevant signals for positioning, competitive scans, or investor conversations—not as audited market data or trading advice. Solo founders in regulated-adjacent categories can use it in early idea and validate phases to frame whether a space is heating up, who is backing whom, and what narrative gaps exist before writing specs. Confidence is moderate until the full SKILL.md body is synced; pair outputs with primary sources and compliance review. It complements build-phase integrations only after you have a clear product hypothesis.651installs25Skill Creatorskill-creator is a Skill Development meta capability from the Anthropic financial-services plugins line, aimed at solo builders who ship with agentic IDEs and want skills that behave like products—not one-off prompts. Prism lists it for builders who need a disciplined way to draft SKILL.md structure, define when the agent should invoke a capability, and refine behavior before promoting a skill to teammates or a public catalog. Use it in Build when you are standing up finance or compliance-adjacent agent tooling, and again in Ship when you harden instructions after review. The ingested readme on this page is license text only; treat triggers and evaluation steps as defined in your local SKILL.md copy after install. Pair it with brainstorming or planning skills when a new skill should follow an approved spec.636installs26Deck RefreshDeck Refresh is an agent skill from the financial-services plugin set that helps solo builders and lean deal teams update presentations when numbers move but the story and design do not. It treats the deck as formatting source of truth and limits work to value substitutions across text runs, tables, and charts. The process starts by clarifying how new data arrives—pasted mappings, Excel columns, or isolated figures—then plans changes before a hard approval gate blocks edits until you sign off. Environment detection separates live PowerPoint add-in edits from chat-side regeneration of affected slides from an upload. Use it when quarterly results, rebased comps, or a single headline metric change must propagate everywhere without a weekend of manual find-and-replace errors.608installs27Ib Check Deckib-check-deck is a reference checker skill that maps everyday finance and startup language to investment-banking-standard phrasing for decks, memos, and transaction materials. Solo and indie founders preparing fundraises, M&A conversations, or advisor-facing documents can invoke it when drafts sound conversational or hype-heavy instead of institution-ready. The skill encodes replacement tables (for example turning cheap valuation into attractive valuation or valuation discount), editorial rules against contractions and exclamation points, and guidance to remove filler like kind of or basically in favor of measurable statements. It does not build models or verify numbers; it tightens how claims are stated so credibility matches the rigor investors expect. Use it while revising narrative sections before sending materials to banks, angels, or strategic buyers, and pair it with your own fact-checking on every percentage and margin cited.604installs