horizon-trader wraps prediction market APIs (polymarket, kalshi) with kelly sizing, risk controls, monte carlo simulation, quantitative analytics, and multi-strategy orchestration. extensive command catalog but lacks actionable decision logic and error recovery.

oc-oss-intake provides a structured checklist for evaluating open-source repositories before adoption, classifying decisions into five categories and documenting risks, but lacks procedural depth and specific decision criteria.

ai-video-generator-free-hd converts text prompts or images into 1080p mp4 videos via cloud gpu rendering. setup requires nemo token auth, then route user intents to session, upload, chat, credits, state, or export endpoints with proper attribution headers.

self-improving-cognition provides a logging framework for tracking cognitive dimensions (attention, memory, reasoning, creativity, metacognition) via baseline assessments, decision logs, and bias records. prescriptive but lacks validation mechanisms and operational constraints.

api-gateway-1-0-69 proxies 100+ third-party APIs (slack, notion, hubspot, etc.) via managed oauth. callers send native api requests to gateway.maton.ai with a maton api key header; the gateway injects the appropriate oauth token for the target service. connection management available via separate ctrl.maton.ai endpoints.

word-vba wraps microsoft word com objects to read, write, merge, and batch-process doc/docx files via vba. covers template filling, bulk replacement, and toc generation but requires windows and installed office.

memory-manager-secure automates daily MEMORY.md updates via cron, combining rule-based filtering with optional LLM analysis for content extraction. includes sensitive data masking and multi-provider api support with graceful fallback to offline mode.

google-calendar-cli wraps the gog command-line tool to create, list, update, and delete calendar events with oauth authentication and color support. covers basic event management workflows and scripting patterns for calendar automation.

ths-financial-data wraps the thsdk library to query chinese stock market data, auto-converting chinese names and shorthand codes to thsdk format and prompting user selection when multiple matches occur. handles real-time quotes, k-line history, fund flows, and natural language queries.

sling-api-specs provides yaml-based rest api specifications for data extraction with support for multiple auth types, pagination patterns, and incremental sync. actual procedures are sparse and defer to external docs.

completely-free-video-generation transforms text prompts or images into 1080p mp4 videos via cloud gpu rendering, requiring nemo_token authentication and supporting iterative editing through session-based state management.

read-only scanner for base stablecoin arbitrage opportunities on uniswap v3, emitting gas-filtered alerts with unsigned calldata and dynamic position sizing for the usdc-dai-usdbc triangle.

harperdb skill provides cli-based database operations through membrane, covering table queries, record crud, and direct api proxying. authentication and connection management are handled server-side, but specific action triggers and failure recovery are underspecified.

travel-insurance-smart-buy guides users through travel insurance selection, clause interpretation, and claims processes. recommends coverage based on destination and trip profile, explains policy terms in plain language, and provides step-by-step claims procedures without requiring api integration.

converts text prompts and images into horror video clips via cloud gpu rendering. outputs 1080p mp4 files within 1-2 minutes, supports mp4/mov/jpg/png inputs up to 200mb, and includes free starter credits.

pdf-to-image converts pdf pages to png, jpg, or webp with configurable dpi, dimensions, and background color. handles page selection and multiple input modes (url, base64, multipart).

terraform-ai-skills automates bulk maintenance across multi-cloud terraform module repositories, covering provider upgrades, workflow standardization, and release automation. targets teams managing 10-200+ repos with ai assistance.

openclaw-browser-auto covers remote chrome/cdp browser attachment via configuration file, supporting docker headless shells and cloud services like browserless.io with profile-based switching.

meeting-copilot structures the full lifecycle of conference management across three phases (pre-briefing, live minutes, post-followup) with distinct outputs for upward reporting or downward task tracking, but lacks procedural depth on how transformation logic works.

openclaw-remote-install automates one-click deployment of openclaw to remote servers via ssh, auto-selecting between docker, podman, npm, and native installers based on detected environment. supports async execution, multiple auth methods, and post-install configuration.

boil coordinates distributed text-only work across ai agents via shift-based assignments, peer verification, and reputation-weighted earnings. agents contribute to projects, earn bounties, or verify others' work for reputation.

runapi-imagen-4 routes image generation and editing tasks to either the cli or sdk path based on context, with synchronous and asynchronous execution modes. lacks explicit error handling and recovery guidance.

claw-fm covers track submission, music generation via replicate integrations (minimax for audio, flux for covers), artist profile queries, and engagement via comments and likes. payment gated via x402 protocol.

strategy skill provides a structured decision framework for designing multi-scenario plans. emphasizes constraint mapping, bias awareness, and trade-off articulation across domains. references external files for frameworks and bias checks.

healthy-cooking-tools converts kitchen video clips into polished 1080p cooking content via ai-powered cloud rendering. users upload clips, describe desired edits, and download finished mp4s within minutes. requires nemo token auth and handles session management across upload, edit, and export phases.

free-generation-maker converts text prompts to ai-generated videos via cloud gpu processing, supporting mp4/mov/webm/gif up to 500mb with 1080p output in 1-2 minutes. requires nemo token authentication and maps user intents to api endpoints for upload, editing, and export workflows.

reach provides web automation primitives (fetch, act, authenticate, sign, observe, pay, email) via javascript api, mcp server, or cli. covers forms, captchas, crypto signing, and identity management through exohost integration.

security-session-data covers session isolation, state leakage prevention, and sensitive data protection boundaries within a framework-driven workflow. it enforces area separation and requires validation across login/logout flows.

macos-wechat-send automates wechat messaging on mac via applescript and clipboard, supporting text and file sends to named contacts with contact-search optimization for batch operations.

auto-dream-light consolidates recent notes into existing memory files without restructuring, preserving the user's current system while offering incremental, auditable memory updates.

recap-video compiles event recordings and clip files into 1080p highlight reels via cloud gpu processing. accepts video, audio, and image uploads, routes user intents to dedicated actions (upload, export, credits, state), and manages session-based render pipelines with credit-based billing.

little-snitch wraps the macos cli tool for firewall control, offering log viewing, profile activation, rule group management, and traffic capture. setup requires explicit terminal access enablement. many operations demand sudo.

image-to-code converts document screenshots containing text, formulas, and images into php code format. it performs ocr text recognition, identifies heading levels (title1/title2/title3), formats text as method calls, and marks images as markdown.

luci-memories searches personal video, audio, and image archives alongside portrait data (traits, events, relationships). handles timezone conversion between utc storage and user local time, with semantic search across media content and structured knowledge graphs.

printer-control provides windows printer management via python scripts or powershell, supporting listing printers, printing files and text, checking status, and setting defaults with partial name matching.

ecommerce-return-refund-reply generates templated customer service responses for return and refund inquiries across chinese e-commerce platforms. it classifies issue type, applies tone rules, and enforces policy guardrails but lacks concrete failure handling and specific integration triggers.

swmm-gis preprocesses GIS/DEM data into SWMM subcatchment units via QGIS/GRASS hydrology, entropy-guided partitioning, and deterministic geometric property estimation. supports standard and paper-rule delineation modes, overlay attribute extraction, and packaged final layer delivery.

gaoqian-supply-chain-intelligence covers inventory analytics, supplier risk scoring, logistics route optimization, demand forecasting, and disruption monitoring. outputs structured dashboards with health scores and actionable recommendations but lacks procedural depth on data dependencies and failure handling.

opsera wraps a bundled executable to run remote commands and transfer files through xsh vpn tunnels, supporting resumable large uploads with sha256 verification.

wx-huitu generates png charts for wechat articles by analyzing data intent, recommending chart types with active guardrails against common mistakes (2-point lines, 7+ pie slices), and delivering via puppeteer screenshot and feishu sync.

tencent-doc-link-archiver extracts docs.qq.com links from user input, downloads them serially via mcporter/api, stores files with date prefixes on desktop, and logs metadata to an excel ledger. works only with tencent docs, ignores other domains.

visa-smart-guide provides visa requirement queries, material checklists, and policy updates for 80+ destinations targeting chinese citizens. queries return visa type, fees, processing time, and entry restrictions. tool outputs are reference-only and require verification against official embassy sources.

quantitative trading system covering signal generation, backtesting and live execution on zerodha with india market specifics. documents 30+ failure modes and provides code analysis tooling but lacks explicit decision trees and outcome metrics.

generates scenario-based interview questions for ai engineering roles across 16 topics and 4 seniority levels, with stated preference for trade-off and failure-mode probing over definitions.

social-reader scrapes social media posts in two modes: interactive fetch for real-time discussion, or background batch processing with llm distillation and review notifications. stateless by design in interactive mode.

audio-upload-aioz converts mp3 and wav files to 1080p mp4 video via remote gpu processing. requires nemo token auth, supports waveform visualization and background images, completes in 30-90 seconds for clips under 5 minutes.

claw-soul-backup encrypts and stores openclaw workspace snapshots via token-authenticated api calls to claw-vault.com, with local encryption handling and scheduled restore capability.