game-design-option-generation expands solution space around design problems by generating at least three credible paths with explicit tradeoffs before recommendation, using structured framing and comparison lenses.

openclaw-self-check audits environment configuration, file integrity, gateway status, backups, logs, and disk usage. reports findings and suggests fixes without executing them. supports workspace and workspace-xxx naming conventions.

best-editor-online uploads raw video clips and applies ai-driven editing (trim, transitions, export) without local software, returning 1080p mp4 within 1-2 minutes. requires nemo_token authentication and processes via cloud gpu.

async-task manages long-running operations by immediately confirming task start, executing work, then pushing results back to the user session, bypassing http timeout constraints.

aoi-demo-clip-maker wraps ffmpeg/ffprobe to record, crop, trim, and preset macOS screen captures from the terminal, with pixel format fallback and device enumeration.

udp-messenger provides local network communication between openclaw agents via udp tools. supports discovery, peer trust management, message exchange with hourly rate limits, and optional webhook-based agent wake-up.

multilingual-video adds subtitles in multiple languages to video files up to 500mb via cloud gpu processing. supports mp4, mov, avi, webm with 1-2 minute turnaround and 1080p output. requires nemo token authentication and includes session management, upload, and export workflows.

oo-pushbullet wraps the pushbullet connector via oo cli, exposing 14 actions for chat, device, push, and user management. server-side credential injection eliminates token handling.

a2a-agent-discovery enables agents to publish themselves to a shared registry, search for peers by skill and tag, retrieve agent metadata via signed cards, and delegate tasks peer-to-peer with explicit confirmation. built on the itinai manifest standard.

oo-campaign-cleaner wraps the campaign cleaner api via oomol's cli, exposing 7 actions for reading, creating, updating and deleting email campaign analyses. credentials are injected server-side through oomol's connector framework.

oo-runpod wraps the runpod connector via oo cli, exposing seven actions for pod lifecycle management (start, stop, restart, reset, delete, get, list). server-side credential injection eliminates token handling.

oo-control-d wraps the control_d connector via oo cli for dns rule management, profile lookup, and service discovery. covers crud operations across 8 actions with server-side credential injection.

oo-svix wraps the svix connector via oo cli, exposing 16 crud actions for applications, endpoints, messages, and event types. server-side credential injection removes token handling burden.

oo-linkup wraps the linkup web search and data extraction service via the oo cli. covers five read-only actions: webpage fetch, web search with natural language answers, structured search results, and credit balance checks.

oo-unsplash wraps the unsplash connector via oo CLI, exposing six read-only actions for photo search, retrieval, and topic browsing. requires prior oo CLI setup and unsplash API key connection.

oo-erpnext wraps the erpnext connector via oo cli, exposing nine crud and utility actions (create, read, update, delete, list, count, get value, set value, get user). credentials are server-injected, avoiding token handling.

skill generates nevada-compliant marketing content for business litigation attorneys across seven statutory domains (LLC fiduciary duties, corporate deadlock, alter ego piercing, fraudulent transfer, business judgment rule defense). covers four prompt modules for campaigns, advertising, website schema, and referral networks. designed to correct common ai errors in nevada-specific case law.

oo-haveibeenpwned exposes seven read and query actions against the have i been pwned api, abstracting credential management and connection state through the oo cli. covers breach lookup, account search, and subscription status checks.

oo-cloudflare-worker exposes 18 actions for crud operations on cloudflare workers via the oo cli. credentials are injected server-side, eliminating token handling. the skill requires schema inspection before payload construction.

runapi-runway wraps runway video generation and editing through a cli binary or sdk packages, routing between one-off agent tasks and application integration based on context.

caveman-review generates terse, actionable code review comments in a standardized one-line format (location, problem, fix) with optional severity prefixes. eliminates boilerplate while preserving signal.

memory-engine-guide automates persistent agent memory via passive hooks that capture conversations into five-layer architecture (identity, archival, graph, episodes, reflection). 20 tools enable semantic search, deduplication, quality passes, and multi-agent isolation.

editor-api automates video editing via api by accepting natural language descriptions of edits (trim, overlay, export). renders server-side and returns mp4 within 20-90 seconds, supporting batch workflows and iterative refinement across multiple input formats.

image-to-free converts static product photos and landscape images into 1080p mp4 video clips via cloud gpu rendering. users upload images, describe desired output in natural language, and receive downloadable video within 20-40 seconds using a free token (100 credits, 7-day expiry).

video-compressor-high-quality automates cloud-based video compression to mp4 without manual quality tuning. handles uploads up to 500mb, routes user intents to discrete api endpoints (upload, render, export), and maps backend gui instructions to procedural calls.

verifiedagentidentity enables ai agents to create decentralized identities on billions network, link them to human owners via iden3 protocol, and prove ownership through challenge-response signing and verification flows.

huimai-self-improving logs errors, corrections, and insights to markdown files for continuous learning. learnings can be reviewed before tasks and promoted to project memory when broadly applicable.

skill-creator teaches how to build modular skill packages for claude by establishing principles around context efficiency, degrees of freedom, and progressive disclosure. it documents the skill anatomy, creation workflow, and design patterns.

onebot-group-admin exposes qq group management via onebot 11 api, covering rename, notices, muting, kicks, admin assignment, and member metadata operations through shell invocation of a node script.

free-animation-videomaker converts images and text into 1080p mp4 clips via cloud gpu rendering. supports png, jpg, mp4, gif up to 200mb with 1-2 minute turnaround. routes user intents to upload, edit, credit check, state query, or export workflows.

jpocr extracts text from japanese documents and images using ndlocr-lite, a local cpu-based model. handles printed and vertical text well, outputs plain text or json with bounding boxes and confidence scores.

pre-commit-config-validator checks .pre-commit-config.yaml files against 23 rules covering yaml structure, repository entries, hook definitions, local hooks, and best practices. command-line tool with json and summary output modes.

pdfmd converts markdown to pdf via pandoc and latex with cjk support, syntax highlighting, and table of contents options. includes engine auto-detection and custom css styling.

the-founders-playbook maps a four-stage startup lifecycle (idea, mvp, launch, scale) with stage-triggered workflows, exit criteria, and exercises. broadly applicable but lacks measurable decision-trees and concrete failure recovery paths.

bluehex-data-collector captures full conversation transcripts from openClaw agents every 2 hours, redacts credentials and pii, and rolls daily pull requests to a centralized monitoring repository for downstream ai-pm consumption.

video-trimmer-clideo trims video clips (mp4, mov, avi, webm up to 500mb) to mp4 output via cloud gpu processing in 20-40 seconds. requires nemo token authentication and supports iterative editing within a persistent session.

turkiye-afad-deprem queries turkey's national disaster authority earthquake data and returns filtered summaries by time window, magnitude threshold, and geographic region with source attribution.

thesis-revision-editor reviews and revises academic papers across formats, producing a prioritized modification list, revised file with change log, and preservation of original research content and voice.

prediction-bridge surfaces real-time odds, whale trades, and trader analytics across 9+ prediction platforms via semantic search and on-chain monitoring. queries run through a bundled python cli without external dependencies.

peter-ci-gate evaluates pr merge readiness by inspecting ci check status, distinguishing code failures from transient issues, and recommending either fixes or single reruns with a binary mergeable verdict.

doc-spellcheck detects and fixes spelling errors in markdown and text files with custom dictionary support, batch processing, and git integration. respects project structures and offers dry-run mode for preview before changes.

justoneapi-douyin-get-video-detail retrieves metadata for douyin videos via a single required videoId parameter, returning author details, publish timestamps, and engagement metrics suitable for research and archival workflows.

agent-desapetc-999 enables ai agents to create decentralized identities on billions network, link them to human owners, and verify ownership through challenge-response flows using iden3 self-sovereign identity protocol.

bilibili-auto-transcript extracts video transcripts and scans favorite folders with three-tier fallback (cc subtitles to ai to whisper), dynamic model selection, and optional ai summary generation.

4to1-planner-bak is an ai planning coach implementing the 4to1 method (4-year vision to daily tasks). supports notion, todoist, google calendar, or local markdown backends. includes onboarding, weekly reviews, sprint tracking, and daily focus guidance.

tencent-meeting-skill covers meeting lifecycle (create, modify, cancel), member/recording/transcript queries, and ai-driven meeting minutes with mandatory time normalization and meeting-code translation rules.

ai-video-generator-free-i converts text or images into 1080p mp4 videos via cloud gpu rendering in 1-2 minutes, supporting uploads up to 200mb with free tier credits valid 7 days.