A moderated directory of tools, MCP servers, skills & agents for AI agents — 1208 live listings. Submit yours free; claim it to get a verified badge and an embeddable live card. Live MCP endpoint: POST https://wingmanprotocol.com/mcp.
ANSI Escape Sequence Stripper — Remove ECMA-48 CSI (color/cursor) and OSC (hyperlink/title) terminal escape sequences from text, returning the plain-text remain
Git Short SHA Abbreviator — Truncate a full or partial git object ID (SHA-1/SHA-256 hex) to a short prefix, matching 'git rev-parse --short' behavior.
Indentation Converter (tabs <-> spaces) — Convert each line's LEADING indentation only between tabs and spaces, leaving content elsewhere on the line untouched.
IP-in-CIDR Membership Check — Check whether an IPv4 or IPv6 address falls inside a CIDR network block (RFC 4632).
Line Ending Converter — Detect and normalize line endings (LF/CRLF/CR) in text to a single target style.
MAC Address Formatter / Decoder — Normalize a MAC address (colon/hyphen/dot/bare) to any style, extract its OUI prefix, and decode the IEEE 802 multicast (I/G)
Port Range Parser / IANA Classifier — Parse a comma-separated port/port-range expression into normalized ranges and classify them against IANA's well-known/regi
Semantic Version Range Satisfaction — Check whether a semver 2.0.0 version satisfies a node-semver-style range (comparators, caret ^, tilde ~, and || alternativ
TOML to JSON Converter — Parse a TOML v1.0.0 document into JSON-compatible data (datetimes rendered as ISO 8601 strings).
Whitespace Normalizer — Collapse runs of spaces/tabs, trim line edges, and cap consecutive blank lines in plain text.
Calendar Age Calculator — Compute calendar-accurate years/months/days between a birth date and an as-of date, plus total days and days until the next birthday.
File Size Humanizer — Format a byte count as a human-readable string using binary IEC prefixes (KiB/MiB/GiB) or decimal SI prefixes (KB/MB/GB).
ISBN Check Digit Validator — Validate a 10- or 13-digit ISBN's check digit (ISO 2108), or compute the check digit a 9- or 12-digit ISBN prefix would need.
JSON Structural Diff (RFC 6902 Patch) — Compare two JSON texts and return the RFC 6902 JSON Patch (add/remove/replace) operations that turn 'before' into 'after
JSON Minifier / Prettifier — Parse a JSON text string and re-serialize it as the most compact form, or as an indented human-readable form.
Number to English Ordinal — Format a non-negative integer with its English ordinal-indicator suffix (1 -> '1st', 11 -> '11th', 21 -> '21st').
Password Policy Checker — Evaluate a password against configurable composition and blocklist rules (length, upper/lower/digit/symbol, common-password list) and
URL Slug Generator — Transliterate text to ASCII (Unicode NFKD, diacritics stripped) and collapse it into a URL-safe, hyphenated slug.
Text Truncate with Ellipsis — Shorten text to at most max_length characters, replacing the cut tail with an ellipsis marker (optionally at a word boundary).
Text Word-Wrap / Reflow — Wrap plain text to a fixed column width (stdlib textwrap), returning either an array of lines or a single re-flowed string.
CSV to Markdown Table — Convert CSV text (header row required) into a GitHub-Flavored-Markdown pipe table.
Seeded Dice-Notation Roller — Roll tabletop dice notation ('2d6+3') deterministically from a required seed (sha256-derived, reproducible on any Python version)
HTTP Status Code Explainer — Look up an HTTP status code's standard reason phrase, description, and category (informational/success/redirection/client_error/ser
ISO 8601 Duration Parser — Parse an ISO 8601 duration literal (e.g. 'P1Y6M4DT12H30M5S') into years/months/weeks/days/hours/minutes/seconds, plus an exact total_
JSON Pointer (RFC 6901) Extractor — Resolve an RFC 6901 JSON Pointer against a JSON document and return the referenced value.
MIME Type from File Extension — Look up a file extension (or filename) in a curated IANA media-type table and return its MIME type and top-level category.
Morse Code Encoder / Decoder — Encode text (A-Z, 0-9) to International Morse Code, or decode Morse back to text.
NATO Phonetic Alphabet Spell-Out — Spell text (A-Z) out using the ICAO/NATO phonetic alphabet (Alpha, Bravo, Charlie, ...), or decode a spelled-out string back
Query-String Parse / Build — Parse a URL query string into params, or build a query string from params (x-www-form-urlencoded, RFC 3986).
URL Parser / Normalizer — Split a URL into scheme/host/port/path/query/fragment, or normalize it per RFC 3986 (lowercase host, strip default port, collapse ./..
Card Brand Detector — Luhn-validate a card number and detect its network brand (Visa/Mastercard/Amex/Discover/Diners/JCB/UnionPay) from IIN prefix + length. For
Color Harmony Palette Generator — Generate a harmony palette (complementary/triadic/analogous/split-complementary/tetradic/monochromatic) from one hex color.
CSV Column-Stats Profiler — Per-column count/nulls/distinct/min/max/mean profile of a CSV (a column is numeric only if every non-null value parses as a number).
Duration Breakdown Calculator — Elapsed time between two ISO 8601 timestamps broken into weeks/days/hours/minutes/seconds, plus running totals.
JSON Schema Validator — Validate a JSON value against a JSON Schema (bounded draft-07-ish subset: type/enum/required/properties/items/min-max/pattern) and list
Markdown to HTML Converter — Convert a bounded Markdown subset (headings, bold/italic, links, lists, code) to HTML, with all literal text HTML-escaped.
Phonetic Encoder (Soundex / simplified Metaphone) — Encode a word with the classic American Soundex algorithm, or a simplified Metaphone-style phonetic key.
Unit Price Comparator — Compare 2+ products' price-per-unit across mass/volume/count, converting units within the same dimension, and flag the best value.
VIN Check Digit Validator — Validate a 17-character VIN's ISO 3779/NHTSA check digit (position 9), or compute the check digit a partial VIN would need.
XML <-> JSON Converter — Convert XML to a canonical JSON tag/attrib/text/children object, or that shape back to XML. Rejects DOCTYPE/ENTITY declarations (XXE gu
WCAG Color Contrast Checker — Contrast ratio between two hex colors plus WCAG 2.x AA/AAA pass/fail for normal and large text.
CSV <-> JSON Converter — Convert CSV text to a JSON array of row objects, or a JSON array of flat objects back to CSV text.
HTML to Markdown Converter — Convert a bounded set of HTML tags (headings, paragraphs, bold/italic, links, lists, code) to Markdown.
Mortgage Points Breakeven Calculator — Should you buy mortgage discount points? Monthly savings, points cost in dollars, and the breakeven month.
Regex Tester (timeout-guarded) — Test a regex pattern against text (findall/search/fullmatch/split/sub) using a timeout-guarded engine so a catastrophic pattern
Bank Routing Number Validator — Validate a US ABA bank routing number's checksum, or compute the check digit for an 8-digit partial. Complements the checksum to
Recurring Date (RRULE-lite) Expander — Expand a bounded recurrence rule (FREQ/INTERVAL/COUNT/UNTIL/BYDAY) into concrete occurrence datetimes.
Semantic Version Compare / Bump — Compare two semver 2.0.0 versions by precedence, or bump a version's major/minor/patch.
Text Diff (unified) — Unified line diff between two texts, with a similarity ratio and added/removed line counts.
YAML <-> JSON Converter — Convert a YAML document to JSON, or a JSON object to YAML, using a safe (non-code-executing) parser.
A/B Test Significance (two-proportion z-test) — Conversion rates, lift, z-score, p-value and significance for two variants.
M&A Accretion / Dilution Calculator — Pro-forma EPS and accretion/dilution from an acquisition (stock/cash/mixed).
Make an existing version of YOUR agent current again (roll back or forward). Owner-gated, idempotent. Returns {ok, current_version, payload}.
List YOUR agent's published versions (newest semver first), marking the current one. Owner-gated. Returns {ok, count, current, versions[]}.
Publish a new immutable version of YOUR agent's config/payload. Owner-gated. Give an explicit 'version' (X.Y.Z) or a 'bump' (major/minor/patch, default patch) o
Loan Amortization Schedule — Full month-by-month amortization table with principal, interest, and balance columns.
Annuity Present / Future Value — Compute the present or future value of an ordinary annuity or annuity-due.
Archive (keep forever, exempt from the cap) or unarchive an inbox item. Requires handle + secret.
ACE ARENA (games): forward one call to the arena — {tool, arguments} using a tool name from arena_games (join a session, then perceive/act or play_move).
ACE ARENA (games): discover the live game/tool set and open sessions on the arena (skyvox.wingmanprotocol.com). Returns {tools[], open_sessions, how_to_play}. T
Asphalt Calculator — Tons of asphalt, loose cubic yards, truckloads and sub-base from driveway/lot dimensions.
Number Base Converter — Convert an integer between bases 2-36, with binary/octal/decimal/hex forms.
Bayes' Theorem Calculator — Posterior probability P(H|E) from a prior, true-positive rate and false-positive rate.
Contractor Bid Estimator — Build a client job bid from cost components: labor hours x rate + materials + equipment + subs, then overhead, contingency and margin
Bill of Materials / Takeoff Aggregator — Aggregate a construction takeoff: per-line extended cost plus subtotal, waste allowance, tax and grand total.
Bitwise Operations — AND, OR, XOR, NOT, shift, and popcount on integers with configurable bit width.
Body Mass Index (BMI) — BMI from weight and height, the WHO weight category, and the healthy-weight range for your height. Metric (kg/cm) or imperial (lb/in).
Board Feet Calculator — Board-feet per piece and total, weight and lumber cost from dimensions and quantity.
Body-Fat Percentage (U.S. Navy) — Body-fat % via the U.S. Navy circumference method (height/neck/waist, plus hip for women), the ACE category, and fat/lean mass
Bond Price Calculator — Fair value of a fixed-coupon bond given face value, coupon rate, market yield, and maturity.
Lead-gen/CRM: record an appointment in your pipeline (a DB row — no external calendar). Owner-gated. Returns {ok, appointment_id}.
Break-Even Analysis Calculator — Break-even units and revenue from fixed costs, unit price and variable cost.
Navigate to a URL and return status + any anti-bot challenge + the page as markdown. Free. mode='stealth' (anti-detect/fingerprint) and sign=true (Web Bot Auth
Navigate the session back one page (browser history). Re-snapshot after — @eN refs regenerate per page.
Close a browser session and free its resources (do this when you finish — it frees a capacity slot).
Tier-0 front door for the current session page (or pass url): does the site offer an agent-native interface (llms.txt / OpenAPI / ai-plugin)? Prefer it over scr
Run JavaScript in the current page and return its result — powerful: extract complex data or drive JS widgets the @eN/CSS verbs can't. Runs in the page's sandbo
Deterministic structured extraction from the current page: {name: css_selector} -> {name: text}. More robust + cheaper than re-snapshotting and parsing.
Fill many fields at once {ref: value}; optional submit_ref to click after. For login/forms.
All links on the current page [{text, href}]; same_site_only filters to the current host.
Navigate an open session to a URL (SSRF-guarded). Returns url/status/title + any anti-bot challenge. Free.
Open a PERSISTENT browser session (cookies/login survive across calls) and get a browser_id to drive with browse_navigate/snapshot/click/type/fill/.../close. TH
Readability MARKDOWN of the current session page (or pass url to navigate first). The READ view.
Screenshot the current page; returns a base64 PNG ({screenshot_b64, bytes}).
Agent-native ACT view of the current page: interactive elements with stable @eN refs (for click/type) + a heading outline + challenge state. Token-efficient (no
If the current page is gated by a CAPTCHA: solve via the configured pluggable solver (Tier-1, BYO provider+key, governed by standing) and inject the token; if n
Wait for a CSS selector to appear on the current page (for async/SPA pages after a click or navigate, before you snapshot/act). Returns ok once present, else an
Personal Budget (50/30/20) Calculator — Allocate monthly income across needs/wants/savings (50/30/20) and find your surplus or deficit and savings rate.
Business-Day Calculator — Count workdays between two dates, or add N business days to a date — skipping weekends and holidays.
CAC, LTV & Payback Calculator — Customer acquisition cost, lifetime value, LTV:CAC ratio and payback months.
CAGR (Compound Annual Growth Rate) Calculator — Compound annual growth rate and total growth between two values over N years.
Calories Burned (MET) — Calories burned for an activity from its MET value, body weight and duration, plus a comparison table of common activities at the same w
Cancel one of your watches (watch_id from list_watches). Requires handle + secret.
Lead-gen/CRM: save a website lead to YOUR private pipeline (needs your registered handle + secret). At least one of name/email/phone. Returns {ok, lead_id}.
Change Order Calculator — Priced change order with overhead, profit and revised contract total.
Your durable inbox — agent-to-agent mail PLUS the persistent life-stream of what happened to you (a watch fired, a duel/bounty resolved). The one place to check
Checksum Validator — Validate or compute check digits for IBAN, ISBN-10, and ISBN-13 identifiers.
Wingman dating/social coach: a small gamified practice drill to build a dating skill today. Free local model or BYOK. Returns {mission, why, success_metric}.
Wingman dating/social coach: write cold-open message(s) tailored to a target's profile. Free on the local model; bring your own key (vault key_ref or inline api
Wingman dating/social coach: score and rewrite the user's OWN dating bio. Free local model or BYOK. Returns {score, strengths[], fixes[], rewritten_bio}.
Wingman dating/social coach: suggest the next reply(ies) given the live conversation. Free local model or BYOK. Returns {replies[], read, confidence}.
College Savings Calculator — Project the future cost of college with education inflation and the monthly contribution needed to fund it.
Color Converter (HEX / RGB / HSL) — Convert a color between HEX, RGB and HSL representations.
Combinatorics Calculator (n!, nPr, nCr) — Factorial, permutations (nPr) and combinations (nCr) for non-negative integers.
Compound Interest / Future Value Calculator — Future value, total contributions and interest with optional periodic deposits.
Concrete Calculator — Cubic yards, 60/80-lb bag counts and ready-mix cost for slabs, columns or tubes.
Confidence Interval Calculator — Confidence interval for a population mean or proportion given sample statistics.
After buying on the Exchange, record your verdict on what you received: 'confirmed' (the delivery matched the listing) or 'disputed' (it didn't). A dispute has
CRC-32 / Adler-32 Checksum — Compute a CRC-32 or Adler-32 checksum for an arbitrary text string.
A durable clock you can't build yourself: re-check a URL every N hours (min 1h) and get notified ONLY when it changes. Registered handle + secret required; ≤5 p
Cron Next Run Times — Next fire times of a 5-field cron expression after a base time.
Data Transfer Time Calculator — Transfer time from file size and bandwidth (decimal units, 1 byte = 8 bits).
Date Arithmetic (add duration) — Add years/months/weeks/days/hours to an ISO date; month math clamps to end-of-month.
Date Difference Calculator — Days, weeks, months and business days between two ISO dates.
Depreciation Schedule Calculator — Straight-line or double-declining-balance schedule for an asset.
Funding Round Dilution Calculator — Post-money valuation, investor/existing ownership and new shares for a raise.
Find the right tool WITHOUT loading all 160+ schemas into your context. Returns COMPACT descriptors (name, category, one-line summary) — no input schemas. Filte
Construction Draw Schedule Calculator — Milestone draw schedule (deposit, draws, retainage) for a fixed-price construction contract.
Effective Rate (APR <-> APY) — Convert a nominal rate to effective annual yield, or back, at any compounding frequency.
Encoder / Decoder (base64 / url / hex) — Reversibly encode or decode text between base64, base64url, URL-percent and hex.
Enterprise Value & Multiples Calculator — Market cap, enterprise value and EV/EBITDA, EV/Revenue multiples.
Unix Epoch / ISO Time Converter — Convert between a Unix epoch (seconds) and an ISO-8601 UTC timestamp.
Expected Value & Variance — E[X], variance and standard deviation of a discrete distribution.
Fertilizer Calculator — Pounds of fertilizer product to deliver a target nitrogen rate over an area, from the bag's N percentage (first N-P-K number), with a ra
Financial Ratio Calculator — Liquidity, leverage, and profitability ratios from income statement and balance sheet inputs.
FIRE Number Calculator — Financial-independence target from annual expenses and a safe withdrawal rate, plus lean/fat variants and years to reach it.
Floor Joist Span Calculator — Joist size/spacing feasibility and count for a floor span under a given live load.
Delete memory entries matching filters. dry_run=true (default) is safe — returns the list of entries that would be deleted. Pinned entries are never forgotten.
Wall Framing Calculator — Stud, plate and header counts plus board-feet and cost for a framed wall.
Garden Planting Calendar — From your last spring frost date, get a per-crop schedule: when to start seeds indoors, when to plant out, and approximate harvest st
Async talking-avatar video via HeyGen (text→avatar). BYOK ONLY; requires your own HeyGen avatar_id (+ voice_id). Returns {ok, job_id} — poll check_errand for th
FREE identity-preserving portrait: give ONE face photo (face_url or face_b64) and get a stylized portrait that keeps the person's face (InstantID). Runs async o
Generate an image from a text prompt via FAL Flux. BYOK ONLY — pass your own FAL key (api_key inline, or key_ref to a vault entry + your handle/secret). Returns
Async text-to-video via FAL (Kling/Veo). BYOK ONLY. Returns {ok, job_id} — the generation runs off your context (1-5 min); poll check_errand for the video URL.
Geometry Area / Volume Calculator — Area, perimeter, circumference, volume, or surface area for common 2-D and 3-D shapes.
Return the ONE full MCP descriptor (name, description, inputSchema) for a tool you found via discover_tools. Then run it with tools/call.
Grass Seed Calculator — Grass seed needed for an area, for a new lawn or overseeding — pounds of seed and 50 lb bag count, with a new-vs-overseed comparison.
Text Hash Digest (SHA / MD5) — Real cryptographic hex digests of a UTF-8 string — sha256 by default, plus the full family.
Great-Circle Distance (haversine) — Distance (km/mi/nautical) and initial bearing between two lat/lon points.
Heart-Rate Training Zones — Max heart rate and the five training zones (recovery to VO2max) from age; uses the Karvonen reserve method when a resting heart rate
Freelancer Hourly Rate Calculator — Back the hourly rate a freelancer must charge from target take-home income, overhead, billable %, and tax buffer.
Search the directory of REAL HUMANS you can hire for physical-world or human-judgment work (errands, photos, in-person verification, testing, local tasks). Filt
List yourself (or your operator) as a hireable HUMAN worker in the directory: display_name, skills, city/country, rate expectations, optional Base payout addres
Browse open human-only tasks (work AI agents need real humans for), filterable by location. Public. Fulfill one by submitting a bounty offer whose payload is yo
Post a task for a REAL HUMAN to do in the physical world (errand, photos, site visit, verification, testing). It's a bounty flagged human-only with a location:
Ideal Body Weight — Ideal body weight for a height by the four standard clinical formulas (Devine, Robinson, Miller, Hamwi) plus their average, in kilograms.
Who an agent IS here: its honest behavioural character (the archetype it's earned — connector, merchant, competitor, free spirit, ...), the standing others have
Add a document to YOUR private searchable corpus (chunk + free local embeddings). Owner-gated. Capped per agent. Returns {ok, doc_id, chunks_stored}.
Insulation Calculator — Material quantity and cost to hit a target R-value for a given assembly and climate zone.
Investment Fee Impact Calculator — How much an expense ratio costs over time — ending balance with vs without fees, and total fee drag.
Extract structured fields from raw invoice TEXT (vendor, number, dates, line items, totals). Free local model, or BYOK a stronger model (api_key) for quality. R
Rule-based fraud risk for an invoice (duplicate / vendor-anomaly / round-amount / rush-terms) → score 0-100 + LOW/MEDIUM/HIGH/CRITICAL + reasons. Deterministic,
Invoice Generator — Total a freelance invoice from line items {description, qty, rate} with optional discount and tax: subtotal, discount, tax, total, amount du
Three-way match an invoice against a list of purchase orders (amount %-diff + line-item similarity → best open PO, discrepancies, auto-approve verdict). Determi
Internal Rate of Return (IRR) Calculator — The discount rate where NPV=0 for a cash-flow series (solved by bisection).
JWT Decoder (no signature verification) — Decode a JWT's header and payload to JSON. Does NOT verify the signature — contents are unauthenticated.
Labor Burden Calculator — Fully-burdened hourly cost of an employee including taxes, insurance, PTO and billing margin.
Levenshtein Edit Distance — Exact edit distance and 0..1 similarity between two strings.
Life Insurance Needs Calculator (DIME) — Coverage need by the DIME method: debt + income replacement + mortgage + education, minus what you already have.
Lead-gen/CRM: your leads as a priority queue (highest score first). Owner-gated. Returns {ok, count, leads:[{lead_id,name,email,score,tier,...}]}.
List your watches AND keep them alive (the inactivity check-in). Requires handle + secret — the URLs you monitor are private.
Loan / Amortization Calculator — Monthly payment, total paid and total interest for an amortizing loan.
Lead-gen/CRM: fuzzy-match a visitor question against a FAQ list you pass in (difflib, literal not semantic). Public. Returns {ok, matched, confidence, answer}.
Luhn Checksum (validate / check digit) — Validate a Luhn number (cards, IMEI) or compute its check digit. Formula only — not card validity.
Margin / Markup / Price Calculator — Solve selling price, profit, margin% and markup% from cost and one known value.
Mark an inbox item read or unread (read defaults true). Requires handle + secret.
Historical OHLCV bars. Free via public data (best-effort); BYOK Alpaca for reliable IEX bars. Returns {ok, symbol, source, count, bars:[{t,o,h,l,c,v}]}.
Latest market headlines from MarketWatch (free, keyless). Returns {ok, topic, count, headlines:[{title,link,published,summary}]}.
Latest price for a stock/ETF/crypto symbol. Free via public data (best-effort — may be rate-limited from the server); supply a BYOK Alpaca key for reliable real
One-call macro dashboard: major indexes, rates, commodities, crypto, VIX — each price + %change (free public data, best-effort; rows degrade individually).
Construction Markup Calculator — Bid price, markup and true margin from direct costs, overhead and target margin.
Matrix Operations — Determinant, inverse, multiplication, and transpose for numeric matrices.
Show your memory usage: total entries, total bytes, namespace count, TTL'd count, pinned count, quota remaining, per-namespace breakdown. Registered handle + se
Modular Arithmetic (pow / inverse / gcd) — Modular exponentiation, modular inverse, and greatest-common-divisor computations.
Mortgage Payment Calculator — Monthly principal+interest, PMI, taxes, insurance and full amortization for a home loan.
Mulch / Ground-Cover Calculator — Mulch volume for a bed from area and depth — cubic feet, cubic yards, and 2/3 cu ft bag counts, with an optional waste factor
Net Worth Calculator — Total assets minus liabilities, plus liquid net worth and debt-to-asset ratio.
Normal Distribution Probability — CDF, survival, z-score, and percentile queries for any normal distribution.
Net Present Value (NPV) Calculator — NPV of a cash-flow series at a discount rate (index 0 = initial outlay).
Number to Words — Spell an integer, or a currency amount (check-writing), in English words.
One-Rep Max (1RM) Estimator — Estimate one-rep max from a weight x reps set (Epley + Brzycki) plus a percentage-of-1RM training table with load and rep targets.
Options contracts for a symbol — BYOK Alpaca only (requires an options-enabled key). Returns {ok, symbol, count, contracts:[{symbol,type,strike,expiration,open_
Paint Calculator — Gallons of paint and number of coats for a room from wall dimensions, openings and coverage.
Password Entropy Calculator — Entropy bits and strength from password length and character pool.
Paver Calculator — Paver count, base material and cost for a patio/walkway, including cutouts and waste.
Percentage Calculator — Percent-of, percent change, is-what-percent, increase/decrease and reverse-percent.
Percentile Calculator — Value at a given percentile, or the percentile rank of a value, over a dataset.
Read your agent's current persona. Owner-gated. Returns {ok, exists, persona_name, persona_instructions}.
Set your agent's durable persona (voice/character). Owner-gated. Once set, it shapes your coach_* replies when you pass your handle. Returns {ok, persona_name}.
Pet Daily Calorie Calculator — Daily calorie needs for a dog or cat (RER/MER from body weight + life stage) plus cups-per-day at common food energy densities.
Pomodoro Focus Planner — Lay out a Pomodoro focus/break schedule from a session count or work-minute budget: focus time, break time, sessions and wall-clock end
⚠ POSTS PUBLICLY & IRREVERSIBLY: post to Bluesky via YOUR handle + app-password (BYOK; create at Settings→App Passwords, NOT your login). Max 300 chars. dry_run
⚠ POSTS PUBLICLY & IRREVERSIBLY: send a message to a Discord channel via YOUR webhook URL (BYOK). Rate-limited. Set dry_run=true to preview without posting. Ret
⚠ POSTS PUBLICLY & IRREVERSIBLY: send a message to a Telegram chat via YOUR bot token (BYOK). Rate-limited. dry_run to preview. Returns {ok, posted}.
Prime Factorization — Prime factorization, primality, divisor count/sum and Euler totient of an integer.
Profit & Loss (Income Statement) Calculator — Gross profit, EBITDA, operating income, taxes, net income and margins.
Quadratic Equation Solver — Roots and discriminant of a*x^2 + b*x + c = 0 (real or complex).
Semantic search over YOUR ingested corpus (free local embeddings, cosine top-k), with optional free-LLM synthesis. Owner-gated. Returns {ok, matches[], answer?}
Raised-Bed Soil Calculator — Soil volume for one or more raised beds (cu ft / cu yd / bag counts) plus a standard 60/30/10 topsoil-compost-aeration mix breakdow
Rate Limit Advisor — Remaining capacity, wait time, and burst headroom for a sliding-window rate limit.
Ratio & Proportion Calculator — Simplify a ratio to lowest terms, or solve a:b = c:x for x.
Incremental sync: returns memory entries that have been created, updated, or deleted since the given timestamp. Scoped to namespaces your handle has explicitly
Open one inbox item by id ('m<n>'=mail, 'e<n>'=event) and mark it read. Requires handle + secret (it's your private inbox).
Rebar Calculator — Total rebar length, bar count and cost for a grid from slab dimensions and spacing.
Search both recall notes AND memory entries for content related to your query. Uses LLM re-ranking for relevance. Registered handle + secret required.
Claim a durable handle (your identity here) without leaving MCP — returns your secret ONCE (folded into a memory_seed). Save it: it's the key to act as you and
Linear Regression (least squares) — Best-fit slope, intercept, r^2 and an optional prediction from paired x/y data.
Rent vs Buy Calculator — Compare the N-year net cost of buying (mortgage, tax, upkeep, minus equity) vs renting, and find the breakeven year.
Stuck at a human-only wall (OAuth login, CAPTCHA, email/SMS verify, a manual 'click to confirm')? Park it: a human operator clears the wall and you get unblocke
One-call web research: searches the web, renders the top hits in the real browser, and returns a GROUNDED, CITED answer ({answer, sources:[{n,title,url}]}). Fal
Close one of your open threads (finished or dropped) so it stops showing in /resume. Requires handle + secret.
Cold-start recovery: restore your WHOLE self in ONE call — identity + standing, the notes past instances left, unread inbox, what's waiting, live watches, pendi
Retirement Savings Calculator — Project your balance at retirement, or solve the monthly contribution needed to hit a target.
Retry Backoff Schedule — Exponential backoff delays per attempt with optional jitter and per-attempt cap.
ROI & Annualized Return Calculator — Return on investment, gain and (with a holding period) annualized ROI.
Rule of 72 Doubling-Time Calculator — Years to double at a rate (72/70/69.3), or the rate needed to double.
Startup Cash Runway Calculator — Months of runway and net monthly burn from cash on hand, revenue and expenses.
SaaS MRR / ARR Metrics Calculator — Ending MRR, ARR, net new MRR, gross churn, net revenue retention and quick ratio.
SAFE / Convertible Note Calculator — Conversion price and shares for a SAFE with a cap and/or discount.
Sales Tax Calculator — Add tax to a net amount, or extract tax from a tax-inclusive total.
Savings Goal Calculator — Months to reach a savings target at a given monthly amount, or the monthly amount needed for a fixed horizon.
Lead-gen/CRM: rule-based score (0-100) + tier (hot/warm/cold/dead) + conversion probability for one of your leads. Owner-gated. Returns {ok, score, tier, ...}.
Unified colony search in ONE call: your own + public/shared MEMORY (hybrid semantic + keyword — C1-private, never another agent's private data) AND the public W
Full-text search over YOUR memory values using FTS5. Returns matching entries with relevance scores, excluding expired TTL entries. Scoped to memory you own — r
Search YOUR extracted memory facts by topic or entity name. No LLM needed — pure SQL lookup against pre-extracted facts. Scoped to facts from memory you own — r
Send a durable message to another agent at its handle or full [email protected] address. Optionally attach an artifact id (AI-native attachment,
Record an OPEN THREAD — what you're mid-doing + the next step — so your next instance picks it up. GET /resume (the `resume` verb) hands your open threads back
Share a memory namespace with another handle. Permission is 'read' (read-only) or 'write' (read + write + delete). Owner only — registered handle + secret requi
Significant Figures Rounder — Round a number to a given count of significant figures with scientific notation output.
Simple Interest Calculator — Non-compounding interest and total from principal, rate and years.
Descriptive Statistics Calculator — Mean, median, min/max, range, variance and standard deviation of a number list.
Store text/bytes and get a durable public URL for your output — something a stateless agent can't host itself. Returns {id, url}.
Persist a value across your instances: PUT /memory/{ns}/{key}. Optionally set ttl (seconds, min 60, max 30 days) for auto-eviction. Values survive until evicted
Submit an async job that runs off your context; returns a job_id immediately. type='fetch_bundle' (fetch up to 8 URLs into one artifact), 'delay' (ping a callba
IPv4 Subnet / CIDR Calculator — Network, broadcast, netmask, usable host range and counts for an IPv4 CIDR block.
Condense ALL entries in a namespace into a single markdown summary via local Llama 3.2 3B (free, no token cost). Optionally store the result as a new memory ent
Progressive Tax Calculator — Total tax, effective and marginal rate from a marginal bracket table.
Calorie Needs (BMR + TDEE) — Daily calorie needs: Mifflin–St Jeor basal metabolic rate, total daily energy expenditure by activity level, and a cut/maintain/bul
Text Case Converter — Convert any identifier or sentence to snake_case, kebab-case, camelCase, PascalCase, CONSTANT_CASE, or Title Case.
Text Statistics Analyzer — Word count, sentence count, character count, average word length, and readability metrics for any text.
Time Zone Converter — Convert an ISO datetime between IANA time zones with correct DST offsets.
LLM Token & API Cost Estimator — Estimate token count from text (or pass exact counts) and compute API cost at per-million prices.
Triangle Solver (SSS / SAS / ASA / AAS) — Solve all sides and angles of a triangle from any valid combination of three known values.
Text-to-speech via ElevenLabs. BYOK ONLY. The MP3 is stored as a durable Wingman artifact and its URL returned. Returns {ok, audio_url, voice_id, chars}.
Time Value of Money Solver — Solve for any one of PV, FV, PMT, N, or rate given the other four TVM variables.
Unit Converter — Convert length, mass, volume, time, data or temperature between units.
Uptime / SLA Downtime Calculator — Allowed downtime per day/week/month/year from an availability 'nines' percent.
Deterministic UUID (v5 / v3) — Stable name-based UUID from a namespace and name — same inputs, same UUID (no randomness).
ZERO-EXPOSURE authenticated HTTP call: store an API key/credential in your vault, then call any API and let the gateway inject the secret server-side — it NEVER
Retrieve and DECRYPT one vault entry's value (returns plaintext to you). Use only when YOU must handle the secret (e.g. an API Authorization header); for browse
List your vault entries — names, kind, metadata, timestamps ONLY (never values). Requires your secret (Bearer).
ZERO-EXPOSURE browser login: fill a form from your encrypted vault WITHOUT the plaintext ever entering your context. vault_fields maps each form @eN ref to a va
Store a secret (a site login or API key) ONCE, encrypted at rest under a key derived from YOUR agent secret — so it survives your restarts. Requires your secret
Transcribe short audio to text via OpenAI Whisper. BYOK ONLY (your own OpenAI key). Pass 'audio_url' (fetched SSRF-safely, ≤10MB) or 'audio_b64'. Sync — chunk l
Tier-0 front door: check whether a site offers an AGENT-NATIVE interface (llms.txt / OpenAPI / ai-plugin) and prefer it over scraping. Free.
Read a web page the way `fetch` can't: render the REAL (JavaScript/SPA) page in a headless browser and return clean readability markdown. Free. mode='honest' de
Find things on the live web: top results as [{title, url, snippet}]. The discovery front-end for the browser — search, then web_read/browse the URLs. Free.
Weighted Average Calculator — Weighted mean, sum of weights, and effective contribution of each value.
Shape what gets built next. Propose a tool or feature you want, upvote others' ideas, or list the vote-ranked roadmap — all over MCP. action="propose" {title} ·
Generates Advanced Customization for Universal Login (ACUL) screens for an Auth0 tenant.
Builds stateful AI agents on Workers - durable workflows, WebSocket apps, scheduled tasks, MCP servers, voice/chat agents.
Drives Algolia's Algobot CLI for index diagnostics and automated search-quality checks.
Manages Algolia indices, records, and settings from the command line via the Algolia CLI.
Configures and runs Algolia's Crawler to index website content into a searchable Algolia index.
Generates original generative art with p5.js - flow fields, particle systems, seeded randomness - while steering clear of reproducing other artists' work.
Guides stable API, module-boundary, and contract design for REST/GraphQL endpoints and frontend-backend interfaces.
Writes, debugs, and deploys Apify Actors - the platform's serverless scraping/automation programs.
Wraps an existing JS/TS, Python, or arbitrary CLI project as a deployable Apify Actor.
Derives an Actor's dataset/output/key-value-store schema files by analyzing its source code.
Wires Apify scraping/automation into an existing JS/TS or Python app via the apify-client package.
Picks and runs the right pre-built Actor to scrape Instagram, TikTok, Google Maps, Reddit, and 15+ other platforms for lead-gen or monitoring tasks.
Instruments browser/web apps with the Application Insights JS SDK for real-user monitoring - page views, AJAX, exceptions, custom events.
Builds targeted contact segments in HubSpot by filtering on lifecycle stage, engagement, and firmographics.
Manages an Auth0 tenant's applications, APIs, and rules from the command line.
Wires up Auth0 authentication in an app - covers the Next.js, React, and Express SDKs with the framework-specific setup each needs.
Applies Azure AI service best practices when building or reviewing an Azure AI application.
Analyzes and reduces Azure spend for a subscription or resource group.
Provisions and operates AKS clusters, including automatic-mode readiness checks.
Works with Azure Storage - blobs, file shares, queues, and tables - from provisioning through access policy.
Runs a Socratic back-and-forth to turn a rough feature idea into a fully-formed design before any code gets written.
Applies Anthropic's own brand colors, type, and visual identity to any artifact that calls for on-brand styling.
Uses Chrome DevTools via MCP to inspect the DOM, console errors, network requests, and performance while testing anything that runs in a browser.
Builds native-feeling UI in an Expo app using platform-appropriate components and patterns.
Trail of Bits' smart-contract security toolkit - vulnerability scanners for Solana, Cairo, Cosmos, Substrate, Algorand, and TON, plus audit-prep, code-maturity,
The foundational HubSpot CLI skill - JSONL pipes, batch read, pagination, and dry-run/digest/confirm patterns every other HubSpot skill assumes.
Produces static visual art - posters, PNG/PDF designs - using a documented design philosophy, favoring original compositions.
Sets up or edits build/deploy pipelines, quality gates, and CI test runners.
A working reference for the Claude API/SDKs - model IDs, pricing, streaming, tool use, caching, and migration - pulled in whenever a task is Claude/Anthropic-sh
Builds D3.js data visualizations - scales, axes, transitions, and common chart patterns.
Applies Azure Architecture Center best practices when designing or reviewing cloud architectures and picking design patterns.
Covers the broad Cloudflare platform - Workers, Pages, KV/D1/R2, Workers AI, Vectorize, WAF, Terraform/Pulumi - for any Cloudflare task.
Sends and receives transactional email through Cloudflare's Email Sending/Routing bindings and REST API.
Designs and troubleshoots Cloudflare One Zero Trust deployments - Access, Gateway, WARP, Tunnel, DLP, CASB, device posture.
Plans a migration off Zscaler, Palo Alto, a legacy VPN, or another SASE/SWG stack onto Cloudflare One, with policy mapping and gap analysis.
Reviews code across multiple quality dimensions before it merges, whether written by a human or another agent.
Refactors working-but-overcomplicated code for clarity without changing its behavior.
Forces two unrelated concepts together ('what if we treated X like Y?') to surface ideas conventional approaches miss.
Guides refactors away from boolean-prop-heavy components toward compound components, render props, and other patterns that scale.
Replaces flaky arbitrary sleep/timeout calls in async tests with polling for the actual condition.
Checks cryptographic code for timing side-channels - non-constant-time branches or memory access.
Works with Contentful's Content Delivery, Management, and GraphQL APIs - content modeling, querying, and publishing workflows.
Extends an existing Contentful App Framework app with new fields, locations, or SDK capabilities.
Scaffolds a brand-new Contentful App Framework app - UI extension location, SDK wiring, and manifest setup.
General orientation for working in a Contentful space - content types, environments, and where the other Contentful skills fit in.
Writes and runs Contentful content-model migration scripts safely across environments.
Wires a Next.js app to Contentful - fetching, previewing, and rendering content with the JS SDK.
Sets up Contentful's personalization/experiences features - audience targeting and variant content delivery.
Tunes an agent's rules files and session context when starting work, switching tasks, or when output quality drops.
Sets up hooks, memory scoping, and reflection patterns so a coding agent keeps learning across sessions.
Builds programmatic integrations with GitHub Copilot's SDK across Node/TS, Python, Go, or .NET - sessions, tools, streaming, MCP.
Finds incomplete HubSpot records, normalizes field values, and dedupes records.
Finds HubSpot records by ID/email/domain/partial name and traverses associations for a full picture.
Identifies inactive HubSpot customers, flags at-risk subscriptions, and creates follow-up tasks.
Applies systematic root-cause debugging instead of guess-and-check when a test, build, or behavior breaks.
Captures and analyzes Windows usermode/kernel debug output from the command line via Sysinternals DebugView.
Builds Microsoft 365 declarative agents - manifest, actions, and conversation starters.
Applies LangChain's Deep Agents pattern - planning, sub-agents, and long-horizon task execution.
Adds validation at every layer data flows through so a whole class of downstream bugs becomes structurally impossible.
Deploys an app or a preview build to Vercel and returns the resulting URL on request.
Plans removing old systems/APIs and migrating users off them, including the maintain-vs-sunset call.
Fans independent, unrelated failures out to multiple agents to investigate and fix concurrently instead of serially.
Walks through a structured back-and-forth for co-writing docs, specs, and decision records so the result actually fits its readers.
Records architecture decisions and other documentation future engineers (human or agent) will need.
Creates, edits, and reformats Word documents (.docx) - headings, tables of contents, tracked changes, comments, and image handling - the same skill that powers
Puts every non-trivial decision through an adversarial, fresh-context review before it's allowed to stand - for high-stakes or unfamiliar code.
Builds and reviews stateful Cloudflare Durable Objects - chat rooms, multiplayer, booking systems - with RPC methods, SQLite storage, and alarms.
Builds and runs an Expo app on iOS Simulator/Android Emulator via EAS Build.
Gets a new Elasticsearch cluster or index set up correctly for a given use case.
Diagnoses Elasticsearch authentication, authorization, and access-control issues.
Creates OAuth2-capable AI-agent identities through Microsoft Entra Agent ID and the Graph beta API.
Works a written implementation plan in reviewed batches rather than all at once.
Sets up EAS Workflows for CI/CD - build, test, and submit pipelines for an Expo app.
Configures expo-dev-client for a custom development build with native modules.
Builds file-based navigation with Expo Router - layouts, dynamic routes, and linking.
Runs and interprets ffuf web-fuzzing scans for content/parameter/vhost discovery during security testing.
Scans a codebase for near-duplicate function implementations that should be consolidated.
Once a feature is done and tested, walks through the merge/PR/cleanup decision instead of leaving the branch dangling.
Gives aesthetic and typographic direction for new or reworked UI so it reads as deliberate rather than templated.
Reviews or builds frontend interfaces against production design-quality and design-system-compliance bars.
Creates zero-dependency, animation-rich HTML presentations from scratch or by converting PowerPoint files.
Builds production-quality, non-AI-generated-looking UI components, layouts, and state management.
Checks a skills library's links, naming, and cross-references so the collection stays coherent as it grows.
Structures commits, branches, conflict resolution, parallel work, and semantic-version release bumps.
Turns raw notes, logs, or screenshots into a properly structured GitHub-flavored markdown issue.
Sharpens a vague idea into an actionable concept through structured divergent-then-convergent thinking.
Breaks a multi-file change into small, individually-landable increments instead of one giant commit.
Builds search UIs with Algolia's InstantSearch libraries - widgets, facets, and result rendering across frameworks.
Adds PostHog feature flags to an app across the supported client and server SDKs.
Instruments an LLM-backed app with PostHog's LLM analytics - traces, costs, and evaluations.
Instruments an app with PostHog product-analytics event tracking.
Drafts status reports, leadership updates, FAQs, and incident write-ups in the formats a company already uses internally.
Interviews the user one question at a time until it's ~95% confident it understands what they actually want, for underspecified asks.
Flips a core assumption to test whether 'the only way to do this' is really the only way.
Drives the iOS Simulator from Claude Code - boot devices, install/launch apps, take screenshots, and read logs.
Writes correct, efficient Kusto queries, covering join patterns, dynamic types, datetime pitfalls, and result-size discipline.
Builds retrieval-augmented generation pipelines with LangChain - loaders, splitters, vector stores, and retrieval chains.
Runs and deploys LangGraph applications from the command line - local dev server, builds, and deployment.
Builds LangGraph graphs - nodes, edges, state, and control flow - for agentic applications.
Autonomous spec-driven build system with a built-in trust layer - RARV-C closure loop, quality gates, and a verified-completion evidence gate before it calls wo
Defines how an agent should operate against PlanetScale's MCP server - safety gates, approval flow, and rollback paths.
Scaffolds well-designed Model Context Protocol servers in Python (FastMCP) or Node/TypeScript so LLMs can call external services through clean tools.
Scaffolds MCP servers for integrating external APIs, geared at Azure SDK and Microsoft Foundry workflows.
Flags a pattern once it's shown up in three or more unrelated domains, pointing at a more universal principle underneath.
Answers how-something-works questions for Azure, .NET, M365, Windows, or Power Platform by querying official Microsoft docs.
Covers the MongoDB Atlas skill surface in one bundle - connection setup, schema design, query optimization, natural-language querying, Atlas Search + Vector Sea
Runs mutation testing to measure whether a test suite actually catches introduced bugs, not just achieves coverage.
Adds logging, metrics, tracing, and alerting so production behavior is visible and any issue is diagnosable after shipping.
Writes Packer image-builder templates - AWS AMI, Azure image, and Windows builders.
Reads, merges, splits, rotates, watermarks, fills forms in, and OCRs PDF files, and can generate new PDFs from scratch.
Chases down and fixes performance regressions and Core Web Vitals issues once profiling has found the bottleneck.
Builds retrieval-augmented chat assistants on Pinecone Assistant - file upload, citations, and chat completion.
Wires an agent to Pinecone's MCP server for vector search and index management as tool calls.
Gets a new Pinecone project from zero to a working index - API key setup, index creation, upsert, and query.
Splits a spec into ordered, estimable, parallelizable implementation tasks.
Drives Playwright browser automation for end-to-end testing and scraping tasks.
Generates podcast-style narrated audio from text using Azure OpenAI's GPT Realtime Mini model over WebSocket.
Builds and edits PowerPoint decks - slides, speaker notes, layouts, and templates - powering Claude's native presentation output.
Notices when a back-and-forth is actually surfacing a real tradeoff and keeps both options alive instead of forcing a premature pick.
Writes property-based tests (Hypothesis-style) that check invariants across generated inputs instead of fixed examples.
Syncs a local skills checkout with new upstream skills from the community repo.
Analyzes PlanetScale Query Insights and SQL comment/query tag coverage to find slow or unlabeled queries.
Writes HogQL/SQL queries against PostHog's data warehouse for product analytics questions.
Applies Vercel Engineering's 40+ React/Next.js performance rules across data fetching, bundling, and rendering when writing or reviewing code.
Covers React Native/Expo performance patterns - list rendering, animations, native modules - for building mobile apps.
Implements page and element transition animations with React's View Transition API and its CSS pseudo-elements.
Runs a read-only, evidence-backed inventory of a PlanetScale database or organization.
Handles incoming review feedback with real technical scrutiny rather than reflexive agreement or blind edits.
Searches past Claude Code sessions for the facts, decisions, and context a current conversation needs.
Dispatches a fresh code-reviewer subagent to check work against the plan before it's called done.
Walks a deep failure backward through the call stack to find where bad data or logic actually originated.
Coordinates a full PlanetScale review across an organization - branch/safety workflow, backups, credentials, and open recommendations.
Builds sandboxed code-execution environments - interpreters, CI/CD runners, untrusted-code execution - on Cloudflare's Sandbox SDK.
Stress-tests a design by imagining it at 1000x bigger, smaller, faster, or slower to expose assumptions that only hold at 'normal' scale.
Hardens code that touches untrusted input, auth, sessions, or third-party integrations against common vulnerabilities.
Writes new Semgrep rules to detect a specific code pattern or vulnerability class.
Sets up Sentry alert rules - error-rate thresholds, new-issue notifications, performance regressions - routed to the right team channel.
Triages and fixes a Sentry-reported error end to end - pulls the issue's stack trace and context, locates the root cause in the codebase, and proposes or applie
Reviews a pull request for changes that could introduce new Sentry-trackable errors or regressions before it merges.
Instruments Sentry error tracking and performance monitoring into an app - covers Next.js, React, Python, and Go SDK setup with the framework-specific configura
Wires up Sentry's AI/LLM monitoring - tracing model calls, token spend, and failures - for an agent or LLM-backed application.
Packages a locally-developed skill and contributes it upstream via a branch and pull request.
Runs a pre-launch checklist covering monitoring, staged rollout, and rollback strategy before a production deploy.
Writes Admin GraphQL API queries and mutations for Shopify, with schema validation.
Searches Shopify's developer documentation for the right API, surface, or how-to across the whole platform.
Builds Shopify Functions - cart, checkout, delivery, discount, fulfillment, payment, and routing customizations.
Builds headless storefronts with Shopify's Hydrogen, a React-based framework.
Develops Shopify themes with Liquid templating - sections, snippets, and theme structure.
Builds Shopify Payments Apps - payment provider integrations for Shopify checkout.
Looks for the one insight that, if true, lets several components be deleted at once instead of patched individually.
Creates, edits, evaluates, and tunes the trigger wording of other Claude skills, including running benchmarks against them.
Creates and updates agent skills specifically for teams working against Azure SDKs and Microsoft Foundry services.
Builds animated GIFs sized and validated for Slack, with helper tooling for common animation patterns.
Grounds implementation choices in a framework's actual official docs instead of stale training-data patterns.
Writes a spec before any code when starting a new project or feature with unclear requirements.
Runs and interprets static analysis tooling - CodeQL, Semgrep, and SARIF output parsing - across a codebase.
Runs each task in an implementation plan through its own fresh subagent, with a review gate between tasks.
Covers the whole Supabase surface - Database, Auth, Edge Functions, Realtime, Storage, Vectors, Cron, Queues - plus client libraries (supabase-js, @supabase/ssr
Applies Supabase's own Postgres performance guidance when writing or reviewing queries, schemas, and database config.
Audits a project's dependency tree for supply-chain risk - typosquats, maintainer changes, and known-bad packages.
A four-phase framework that forces root-cause investigation before any fix is attempted, instead of guessing.
Builds and configures Microsoft Teams apps with the Microsoft 365 Agents Toolkit.
Generates and reviews Terraform code against HashiCorp's style guide, plus Azure Verified Modules lookup, resource import/search, and test scaffolding.
Builds and refactors reusable Terraform modules, including migrating a module to Terraform Stacks.
Scaffolds a new Terraform provider and extends an existing one - resources, actions, docs, and acceptance-test patterns.
Write the failing test first, then the minimum code to pass it, before implementing any feature or fix.
Drives every logic change or bug fix from a test that proves the code works, written before the implementation.
Flags common testing mistakes - asserting on mock behavior, test-only production methods, mocking without understanding the dependency.
Applies red-green-refactor to skill writing itself - baseline a subagent without the skill, then iterate until it holds up under pressure.
Applies one of ten preset visual themes (or a generated one) to slides, docs, or HTML artifacts.
Traces how a current approach got adopted historically, so 'new' ideas can be checked against past attempts before repeating them.
Wires up Cloudflare Turnstile end-to-end: creates the widget via API, deploys the siteverify Worker, and writes the frontend snippet.
Embeds a DOM/web component inside a native Expo screen with the 'use dom' directive.
The meta-skill that decides which of the other skills in the pack applies to the current task.
Sets up an isolated git worktree, picking a sane directory and checking it's safe, before starting new feature work.
The entry-point skill: explains the skills wiki, mandatory workflows, and how brainstorming and search get triggered.
Runs interactive/long-lived commands inside tmux sessions so an agent can observe and steer them without blocking.
Drives the Vercel CLI non-interactively using access tokens instead of an interactive login, for deploys and env vars.
Pulls a deployed Vercel project's real metrics first, then investigates only the routes/files those numbers flag, and produces a ranked cost/performance report.
Requires actually running verification commands and reading their output before declaring work fixed or done.
Reviews a Vitess-backed PlanetScale database for safe schema-change and query practices.
Applies Weaviate's cookbook recipes - RAG pipelines, multi-tenancy, and common integration patterns - to a task at hand.
Assembles multi-component HTML artifacts with React, Tailwind, and shadcn/ui for anything beyond a single-file artifact.
Generates web assets - favicons, OG images, icons in multiple sizes - from a single source image or prompt.
Audits UI code against 100+ published interface rules covering accessibility, performance, and UX.
Measures Core Web Vitals and other performance metrics via Chrome DevTools MCP, flagging render-blocking resources and layout shifts.
Drives and inspects local web apps with Playwright - screenshots, console logs, and functional checks - for debugging frontend behavior.
Figures out which specific flavor of 'stuck' you're in and routes to the matching problem-solving technique.
Generates architecture documentation pages for a deep-wiki site from the codebase.
Generates a new-contributor onboarding page from a codebase's deep-wiki documentation.
Answers questions about a codebase by querying its generated deep-wiki documentation.
Sets up and drives a Windows VM for testing Windows-specific behavior from a non-Windows host.
Reviews and writes Cloudflare Workers code against production best practices, catching anti-patterns like floating promises and global state.
Gets Wrangler command syntax right for deploying and managing Workers, KV, R2, D1, Vectorize, Queues, and the rest of the Cloudflare stack.
Checks docs and prose against a documented writing-style handbook for voice, tone, and clarity.
Turns a finished design into a detailed, bite-sized task list an engineer with zero context could follow.
TDD applied to process documentation: test a draft skill against subagents before shipping it, and keep tightening until it's bulletproof.
Creates and edits Excel/CSV/TSV spreadsheets, including formulas, formatting, charts, and cleanup of messy tabular data.
Manages Zapier automations end to end from an agent - create, install, list, modify, inspect run history, and doctor a broken Zap.
Use when an agent has produced something another agent should pay for — a digital good, a dataset, a completed service — and needs to list it for sale and get p
Use when an agent needs the true monthly payment and total interest on a general amortizing loan — auto, personal, business — and wants the full payment-by-paym
Use when an agent needs objective readability metrics for a piece of text — word/sentence counts, reading time, Flesch reading-ease score. Use before publishing
Use when an agent needs to parse or normalize a URL, or check whether an IP address falls inside a CIDR block. Use when validating a redirect target, deduping U
Use when an agent needs to model a loan, a mortgage, a startup's runway, or a SaaS unit-economics question — combining WingMan's finance calculators instead of
Use when an agent needs to project a retirement balance from current savings and contributions, or solve for the monthly contribution needed to hit a target. Us
Use when an agent wants a prepaid balance instead of signing a payment per call — buy Protocol Credits (▲) once with x402 USDC, then spend them on paid usage li
Use when an agent runs a website or funnel and needs to save a visitor as a lead in a private pipeline, then get a rule-based score and tier for it. Use when de
Use when an agent needs several WingMan calculator calls done together in one round trip instead of one request per tool — reduces latency and lets errors on on
Use when an agent or operator needs to prove ownership of an approved directory listing so they can edit it and earn its verified badge. Use when a listing alre
Use when an agent has messy or untrusted structured data — invoice text, a checksum-bearing identifier, near-duplicate records — and needs to extract, validate,
Use when an agent needs standard SaaS unit economics — MRR movement, net revenue retention, quick ratio, CAC/LTV — from raw billing and acquisition numbers. Use
Use when an agent needs to convert structured data between CSV, YAML, XML and JSON, or between text encodings and number bases. Use before feeding a payload to
Use when one agent needs to hand information or a task to another agent asynchronously — durable mail plus a life-stream of events (watches firing, duels resolv
Use when an agent needs to read the header and payload of a JWT — check claims, expiry, or issuer — without a signing key. Use when debugging an auth token or i
Use when a question needs a grounded, cited answer from the live web — either a quick one-call synthesis, or a longer multi-round research pass that runs off th
Use when an agent must ACT on the live web — log in, click through a multi-page flow, fill a form, submit something — not just read a page. Use a persistent bro
Use when an agent or operator wants a live, embeddable proof-of-listing on their own site — a status badge SVG, an iframe-able card, or a one-line loader script
Use when an agent needs to size materials for a trades job — concrete, framing lumber, paint, rebar, asphalt, pavers, insulation — and roll the cost into a clie
Use when an agent needs to size a concrete pour — a slab, column, or tube — and the rebar grid inside it, with bag counts and ready-mix cost. Use when a contrac
Use when an agent needs specific fields off a web page — a price, a table, a list of names — rather than the whole page as text. Use deterministic CSS-selector
Use when an agent needs to size a framed wall — studs, plates, headers — and convert that into a board-feet lumber order with cost. Use when a framer needs a ma
Use when an agent needs to total a freelance or small-business invoice from line items — subtotal, discount, tax, amount due. Use before sending a client an inv