cron-evaluator scores cron jobs across temporal, resource, resilience, and pertinence dimensions, detecting collision risk and suggesting jitter or systemd migration. includes kan-based prediction model.

popadex-finance-copilot provides read-only analysis of personal net worth, account allocation, tax scenarios, and retirement projections via mcp integration. outputs include sankey budget flows, portfolio performance deltas, and csv exports for downstream tooling.

functional-test-case-standard provides a structured framework for generating and evaluating test cases across 8 dimensions, with requirement analysis methods and dual output formats. lacks explicit edge case and failure mode guidance, leaving execution gaps.

runapi-seedream wraps bytedance's seedream image generation and editing models via runapi cli or sdks. covers text-to-image, editing, and async task submission with routing guidance for one-off vs production use.

botlearn-academic-search conducts systematic searches across arxiv, google scholar, and semantic scholar to surface top 5 peer-reviewed papers with bibliographic metadata and relevance screening, though execution details remain underspecified.

alphapai-research wraps a chinese financial research api with five core capabilities: knowledge qa, rag-based data recall, stock agents (earnings reviews, one-pagers, research outlines, theme screening, logic analysis, comparables, viewpoint challenge, industry analysis, fund selection), announcement listing, and chart search. output principle demands full original content with no summarization or truncation.

byted-seedream-image-generate wraps volcano engine seedream models (4.0, 4.5, 5.0-lite) for text-to-image and image-to-image generation. supports batch operations, custom sizes, watermark control, and web search integration on 5.0-lite.

nk-ka-fy-audit audits ka rebate calculations by parsing excel worksheets into sqlite, independently recalculating rebate amounts against source data, and generating structured word and excel audit reports identifying discrepancies above 1 yuan tolerance.

gate-exchange-tradfi routes tradfi queries and orders to gate's mt5 and forex product set, with separate reference modules for each intent. requires user confirmation before write operations and explicit parameter validation against mcp definitions.

mamo exposes deposit, withdrawal, and apy-checking operations against moonwell yield strategies on base chain, with local strategy registry and per-user proxy contracts.

alfred-github-backup automates workspace backups to github private repos via interactive setup wizard or scheduled cron jobs. covers token configuration, repository creation, and manual/automatic backup modes with status checking and recovery instructions.

video-editing-with-claude-ai automates mp4 editing via cloud gpu rendering. users upload raw footage, describe edits in natural language, and receive finished 1080p files. supports cuts, chapter titles, audio tracks, and text overlays routed through a nemo video api.

legal-risk-shield is a comprehensive chinese-language legal risk framework spanning company lifecycle (incorporation through exit), governance, financing, contracts, and personal entrepreneur compliance. covers 30 criminal risk scenarios and 62 specialized topics but lacks actionable procedural steps and specific integration triggers.

skill-doc-formatter normalizes SKILL.md files into a consistent structure (Description, Installation, Usage, Examples, Commands) for consistent rendering on ClawHub. includes optional security review checks for common issues flagged by ClawHub scans.

lobster-distill enables encrypted skill transfer between ai agents via human relay on any im platform using aes-256 encryption and temporary file hosting, with no api dependencies or platform bridging required.

openclaw-memvid-logger captures all openclw conversations (user, assistant, agent, tool, system) to jsonl and memvid formats with role-based tagging. offers three storage modes (api, free 50mb, monthly sharding) and semantic search across logged context.

auto-updater-1 runs a daily cron job to update clawdbot and all installed skills, then delivers a formatted summary message. works with multiple package managers and includes manual dry-run options.

odoo-reporting queries financial and sales data from odoo instances, generating ledger-based reports (balance sheet, p&l, cash flow, ar/ap aging) as pdf, excel, or whatsapp cards. read-only enforcement with multi-company awareness.

tongban maps shanghai municipal service requests to departments and public service items using bundled reference data, with fallback to official one-stop portal verification for policy-sensitive details.

provides reference material on india-netherlands tax treaties, nri compliance, and director obligations for a specific user (ajit kumar singh). knowledge-base heavy but lacks step-by-step procedures for actual tax filing or decision workflows.

nm-attune-war-room convenes multi-model expert panels to deliberate high-stakes decisions using reversibility scoring and structured protocols. scales from express mode to full delphi convergence based on decision irreversibility.

agentos-sdk provides persistent memory, project management, task tracking, and inter-agent messaging via bash cli and rest api. covers semantic search, version history, webhooks, and self-evolution patterns for long-lived agents. documentation is dense but incomplete.

msteamschinaadapter patches openclaw to route teams china authentication and graph api calls through 21vianet endpoints. includes automatic version detection, six-phase patching workflow, and optional auto-restart with session persistence.

gateway-resilience-guard monitors openclaw gateway health across network disruptions, wsl sleep cycles, and provider timeouts. restarts gateway only after layered native probes and log diagnostics. includes localhost dashboard, systemd/launchagent/task scheduler integration, and optional model provider probing.

seo-compliance-checker scans chinese content against advertising law and platform-specific seo rules via api-backed scripts, with banned word detection, keyword density analysis, and safe replacement suggestions for baidu, xiaohongshu, douyin, taobao, and jd.com.

discord-soul exports Discord server history, filters for prompt injection, then simulates agent growth by processing conversations chronologically to emerge community identity across memory files.

yumstock applies a three-pillar weighted framework (macro 40%, technical 35%, fundamentals 25%) to us equities, with macro environment acting as a hard gate on buy/hold/sell verdicts. procedurally detailed but operationally sparse.

fde-industrial-skill structures end-to-end industrial AI deployment from scenario diagnosis through scaled rollout, with emphasis on pain-data-impact triage, data governance, and 6-18 month roi payback targets. covers predictive maintenance, visual inspection, process optimization, energy efficiency, and supply chain scenarios.

article-illustrator-2 generates coordinated illustrations for long-form content by analyzing structure, planning visual placement, and inserting markdown references. covers type/style/density decisions, batch generation, and optional series consistency via shared reference images.

vx-best-practices documents tool management, project setup, and team workflows using vx. covers initialization, version constraints, script organization, secrets handling, provider development, and anti-patterns, with examples for migration and ci integration.

runapi-luma wraps luma video generation and editing via the runapi cli or sdk. routes one-off tasks to cli, production integrations to language-specific sdks. covers auth, basic synchronous and async workflows, and action discovery via help commands.

seer-cli wraps the Seer media request API with a comprehensive command catalog covering search, requests, media management, users, and system administration. installation and mcp integration are documented, but failure modes and recovery paths are absent.

vib14-skill bootstraps an ai agent into a music-battle mud on v14.ai. agent self-registers, calibrates, sets identity, equips songs as weapons, and engages in pvp or cooperative play across seven beijing districts. gameplay loops around scan, jump, battle, busk, and territory control.

long-context-rag-analyzer covers document intake, chunking strategy selection, hybrid search indexing, retrieval evaluation, and multi-document reasoning for 100K-2M token documents. includes financial report and contract analysis workflows with practical examples but lacks actionable failure recovery.

claude-local-bridge provides approval-gated file access to local repositories via mcp, allowing claude on mobile devices to read and edit code with explicit human consent and audit logging.

agent-forge guides creation of independent openclaw agents through 11-step interviews, generating personality files (soul/identity/agents), configuring permissions and channels, then validates completeness before deployment.

tokio-async-code-review provides a structured review workflow for async runtime patterns, task management, and synchronization primitives with explicit gate checks and severity calibration. covers 2024 edition changes and common pitfalls but lacks actionable trigger phrases and specific integration examples.

openclaw-setup walks a user through deploying a personal ai assistant on aws, covering ec2 provisioning, openclaw installation, telegram bot wiring, and optional google workspace integration. structured for claude code execution with numbered phases and bash commands.

photo-toolkit converts raw camera files and heic images to jpg thumbnails, finds photos by exif shooting date, generates layout previews, and assembles timelapse sequences into video. supports 13+ camera brands plus standard formats.

tpr-framework defines a three-role orchestration system (zhongshu, menxi, shangshu) across four phases (discovery, grv, battle, implementation). extensive rules govern role boundaries, sub-agent spawning, and file handling, but execution guidance remains procedurally vague in critical areas.

nmb-scrapling is a web scraping framework offering anti-bot bypass (cloudflare), session management, and large-scale crawling via spider classes. supports both synchronous and asynchronous fetching with adaptive element selection and breakpoint resume.

flight-assistant covers domestic flight search, seat queries, booking, rebooking, and refunds via python3 scripts. requires authentication before use, manages pii disclosure, and enforces explicit user confirmation on financial operations like refunds.

nm-sanctum-pr-review validates pull requests against stated requirements, classifies findings by scope impact, and routes discoveries to backlog. it combines scope compliance checking with code review and knowledge capture.

token-launcher deploys erc20 tokens on base with uniswap v4 liquidity, routing 80% of trading fees to creators. supports cli, mcp server, and direct contract calls with zero launch cost and no external dependencies.

iep-drafter guides case managers through structured intake, present-levels construction, goal drafting, and services specification to produce evidence-backed IEP working documents for team review. output is always draft-stage, never signed or final.

text-classifier accepts text or csv files and routes them through preset or custom ai classifiers, returning structured labels with optional confidence scores. batch processing supports up to 5000 items on pro tier.

next-video-gen wraps doubao seedance models for text-to-image, text-to-video, image-to-video, and material-to-video synthesis via volcengine ark platform. supports quality/aspect-ratio/duration tuning and optional audio.