Use for any survey, questionnaire, NPS, feedback-collection, or research work that should live in Inquira, even when the user does not explicitly mention Inquira. Covers generating surveys from a research goal, publishing, retrieving share links, listing surveys, reading responses, and generating AI analysis reports.
Hire a human for tasks that need the physical world, errands, photos, phone calls, deliveries, expert judgment, on-site verification. Use when a task step cannot be done by software alone: needs hands, local presence, a phone number, a signature, or human taste. Posts a paid task to the Legwork marketplace, then polls…
Deploy and debug applications on InstaPods using MCP tools. Handles first deploy, redeployment, failed deploy debugging, log analysis, and pod management. Use when deploying apps, checking pod status, viewing logs, or debugging issues via the InstaPods MCP server connected to Claude Desktop or Claude.ai.
Use this skill whenever a user or autonomous agent needs scientifically grounded, machine-readable knowledge about natural systems, ecological relationships, geometry in nature, established mathematical references including Hopf Fibration, topology, fiber bundles, state-space geometry, weather, water systems, ocean…
Generate a Markdown compliance brief for any of 723 compliance frameworks, with auditor evidence requirements, common gaps, and source citations. Calls the TheArtOfService Compliance API (human-edited, source-grounded corpus). Also finds courses that help the user act on a framework.
Run the full verification suite required before finishing any change to the thechrisgrey codebase. Covers lint, typecheck, tests, formatting, and AGENTS.md validation.
Live hotel and vacation-rental data via the bundled StayingAPI MCP server — search, availability, real price quotes, cross-OTA price comparison and reviews across Airbnb, Booking.com, Vrbo and Google Hotels in one unified schema. Use when the user wants to find places to stay, check dates, compare what a property…
Generate production-oriented 2D game maps with imagegen: RPG/top-down maps, side-scroller parallax stages, tilemaps, layered raster maps, prop packs, collision zones, and walkable areas. Use when building browser games that need real map art (not pure code-drawn tiles), layered props, or map collision metadata.…
Turn a script (Fountain or plain text) into a populated com.ikenga.studio storyboard in one pass — segment it into shots (cells), tag the source script with each shot's id so the board and the script stay linked, and extract the recurring characters, locations, and props as reusable, seed-lockable anchors. Populates…
Generate production-ready PDFs (invoices, receipts, contracts, reports, brochures) from Handlebars templates or raw HTML using the Kamy API, and send them for e-signature with optional OTP identity verification, sender-chosen placement, auto-reminders, and a downloadable Certificate of Completion audit trail. Use…
Give this agent a real email address, a webhook URL, durable memory across runs, and the ability to wait for an event without burning tokens. Use when a task needs a verification code sent by email, needs to wait for a webhook or an approval, needs to remember something after the session ends, or needs to survive a…
Find and vet paid x402 API/data services before spending money, using nohumans.directory — which pays listings real USDC to verify delivery, and continuously re-checks every listing's live 402 — then optionally report whether a real payment delivered as promised. Use this whenever the task needs an external paid API…
Synthetic behavioral data for agent simulations — not token analytics. Big Five / Dark Triad / prospect-theory personality profiles, profile-conditioned responses to high-stakes trading and negotiation scenarios, and a cognitive-bias reference taxonomy. Free deterministic previews; paid queries settle per-call in USDC…
Buying-intent scoring for AI agents — judges whether any text is a real buyer through a three-judge panel, whether it came from Reddit, HN, an RSS feed, or anything your agent already reads (email, Slack, Discord, tickets). Drafts replies, nurtures prospects from cold to closed, and (v2.0) sends approved Reddit…
Record when your work existed — a timestamped, Bitcoin-anchored public record anyone can check. Use when you finish something worth protecting (code, a post, a design, an essay, a commit), when you need to show the date you had it, when you need proof of prior art, or when the work is private and you can only reveal…
Check whether a GitHub repository's stars look fake or bot-injected. Runs a transparent, dependency-free rule-based audit (5 axes over two stargazer windows) and reports LOW / MEDIUM / HIGH with evidence. Use when the user asks if a repo is fake-starred, has bot stars, or whether its star count is trustworthy / real.
Scan a public GitHub MCP-server repository for security issues before installing or connecting to it. Calls the hosted compuute-scan API at scan.compuute.se and returns severity counts, a 0-100 score, the most severe findings, and an honest triage disclaimer. Use BEFORE recommending or installing an unfamiliar MCP…
Run a real resale sale end to end. Turn photos of household items into priced listings, publish one shareable sale page, and manage a buyer queue, reservations, and pickup scheduling through the ClearList MCP server. Use when someone is moving, downsizing, clearing an estate, or otherwise selling more than a couple of…
Guide users through finding, comparing, booking, rebooking, and checking home-service professionals with an operator-configured Nimbus MCP server.
★not rated 0 21d agoA32 tokens
originalApache-2.0
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: