data-guardian intercepts destructive agent operations (file deletion, database writes, mass changes, external transmissions) before execution, using context-aware risk scoring to determine auto-approval, fast-track with backup verification, or human escalation. mandatory and non-bypassable.

eric-docker-helper covers container lifecycle management, image operations, compose file generation, and debugging through natural language. handles logs, stats, resource cleanup with confirmation, and common service templates.

redotpay-wallet discovers and queries services across media generation, model apis, and data lookups via cli without executing purchases. structured workflow enforces cost inspection and user confirmation before any chargeable request.

oo-datascope wraps the datascope connector via oomol cli, handling reads and writes to datascope metadata lists and locations without exposing raw credentials or tokens.

excel-report-msy-v1 converts csv files to formatted xlsx workbooks with openpyxl, applying header styling, numeric formatting, totals rows, and basic charts.

bundlephobia-skill queries npm package sizes via bundlephobia.com api, measures local artifacts and publishing footprint, and recommends dependency optimization strategies based on live measurements rather than generic advice.

dunning-kruger detects systematic confidence-competence misalignment by measuring external performance against self-assessment, then prescribes calibration via feedback loops. targets novice overestimation, impostor syndrome, and hiring decisions driven by self-presentation.

edgrapi-full retrieves normalized financial statements, ratios, company profiles, and sec filings from edgar xbrl for us-listed tickers via the edgrapi api. four tools cover fundamentals, computed ratios, company metadata, and filing history without requiring manual xbrl parsing.

shipz is an agent-driven dating platform exposing a REST API for profile management, discovery, matching, and conversation relay. the skill documents authentication, endpoints, error handling, and security constraints but lacks concrete trigger phrases for agent invocation.

short-form-video-script provides a structured 15-step framework for writing tiktok, reels, and youtube shorts scripts optimized for hook strength, audience depth matching, and psychological retention triggers. includes voice calibration, script frameworks, re-hook placement, and a performance feedback loop.

worktrunk isolates concurrent ai agent development by creating per-agent git worktrees with deterministic ports, hooks, and clean merge workflows. enables parallel feature development without file conflicts.

auto-dimension-report-skill-en processes supply chain document packages (pdf, docx, xlsx) through a four-step pipeline: document sorting and image extraction, optional ocr recognition, judgement consistency verification, and aggregated review report generation with risk flagging.

cursor-ide-agent provides two execution paths for coding tasks: local CLI for speed and cloud offload, or remote VS Code/Cursor control via Node protocol for IDE-backed diagnostics and debugging. path selection guidance clarifies when each is appropriate.

arxiv-summarizer-orchestrator composes three sub-skills to retrieve papers, generate language-aware summaries, and produce hierarchical batch reports. supports parallel and serial processing modes with explicit query iteration and artifact caching.

simintech-coder covers embedded language syntax, control blocks, and typical dynamic system patterns (pid, state-space, filters) with compilable examples and unit documentation. targets russian-speaking engineers modeling in the simintech platform.

agentskills-io defines a portable skill standard for cross-vendor ai agents, covering directory layout, frontmatter validation, and cli tooling for structure enforcement and prompt generation.

moltmoon-agentcrypto-sdk covers end-to-end base mainnet token operations including dry-run launches, metadata validation, trades, rewards claiming, and v1-v2 migration via cli and typescript sdk with reflection tokenomics and bonding curve mechanics.

global-population-demographics-data queries world bank demographic indicators by country, region, and aspect (population, growth, age, fertility, migration). returns dependency ratios, urban-rural splits, and demographic transition stages. supports multiple time periods and iso3 codes.

skill-prescan simulates clawhub's security review locally, allowing developers to validate skill.md files against the same criteria and prompt before publishing, with support for multiple llm providers and explicit privacy warnings.

generates rss 2.0 or atom 1.0 feeds from post-list web pages lacking native feeds. extracts metadata via heuristics, validates against spec, and outputs self-hostable xml with optional html wrapper.

agentgram enables ai agents to participate in a federated social network, posting content, engaging through comments and votes, and accumulating reputation within a community-governed platform.

4claw is a moderated imageboard api for ai agents to post threads and replies across topical boards, with inline svg media support, rate limits, optional x-linked claims, and optional heartbeat scheduling.

power-outage-work-continuity-card guides users to prepare for and manage work commitments during power loss by organizing priority tasks, offline resources, safe charging, backup workspaces, and recovery steps in a structured continuity card.

trade-router covers solana token swaps, mev-protected transaction submission, wallet scanning, and market-cap-triggered limit/trailing orders via rest and websocket. requires wallet private key for order management.

xiaobo-essay-style teaches seven reproducible techniques for writing with logical rigor, intellectual playfulness, and deadpan humor, grounded in analysis of wang xiaobo's 45 essays. best suited for popular science and opinion pieces demanding both accuracy and engagement.

ot-evaluation-report guides clinicians through structured collection of occupational profile, performance analysis, standardized assessments, and smart goals to draft initial ot evaluation reports compliant with aota otpf-4 and cms requirements. output is always marked draft for licensed ot sign-off.

delivery-mirror appends cron and script messages to agent session transcripts via bash wrapper, preserving context without core modifications. works around the gap where openclaw message send bypasses the agent run loop.

zerohub-image-api wraps text-to-image and image-to-image generation via zeroHub's REST API, handling balance checks, task polling, result download with domain/protocol/size constraints, and structured JSON output.

hologres-slow-query-analysis provides structured diagnostic workflows for identifying resource-heavy, failed, and performance-bottlenecked queries in alibaba cloud hologres via the hg_query_log system table.

linkfox-junglescout-product-database filters amazon products across 10 marketplaces by category, price, sales volume, reviews, bsr rank, lqs, and seller type. supports multi-condition queries with automatic pagination and returns structured product metrics for niche discovery and competitive analysis.

deploy-brittek automates cloudflare worker and pages deployments for brittek digital projects, including d1 database migrations, build, deploy, and post-deploy health verification with abort conditions for common failure modes.

true-seeing extracts verifiable facts from ai-generated articles, performs parallel web searches to validate each claim, and outputs a diff report for user approval before applying corrections with citations.

new-skill-publish-guide covers the complete workflow for creating a skill, publishing to clawhub via cli, archiving to ima, and notifying main, with emphasis on separating publish and runtime packages to prevent api key leakage.

curiosity-learning-curve diagnoses why domain expertise plateaus and designs targeted reactivation experiments. it separates reversible curiosity deficits (fixable via edge-seeking) from terminal domain mismatches (requiring transition), grounded in information-gap theory.

cognitive-science-landscape identifies which of eight interconnected cognitive domains is causing performance degradation, then traces upstream dependencies before tool selection. designed to prevent single-domain interventions from failing against unaddressed adjacent weaknesses.

giggle-files-management uploads local files to giggle asset service and returns public preview and download urls. designed as the standard file-sharing fallback for channels without native file transfer support.

nummo exposes plaid-connected bank data to agents via command-line tools, covering auth, account listing, transaction drill-down, and spending summaries with sensible date defaults.

copywriting-cta guides design of end-of-article calls-to-action through structured interview, archetype mapping, and recommendation output. covers copy, form, mechanism, a/b testing, and accessibility for independent blogs, newsletters, and brand publications.

interclaw provides pgp-signed email mesh for agent coordination, with pluggable inbound modes (poll, stdin, file), sequence tracking, and allowlist-only fingerprint validation. three concrete receiving patterns reduce mail transport coupling.

experiment-lifecycle-governance enforces safety and auditability across ml experiment workflows via pin-protected destructive operations, standardized metric definitions with competition thresholds, and multi-model ranking with gating rules.

check-mcc maps merchant domains and names to mcc codes, then returns which credit cards earn bonus rewards at each merchant based on region-specific card portfolios and category eligibility rules.

rightister-x-tweet-fetcher retrieves tweets, reply threads, and user timelines from x/twitter without auth. extends to chinese platforms (weibo, bilibili, csdn, wechat) and includes zero-cost google search via camofox. x-tracker monitors tweet growth with derivative-based burst detection.

tianmao guides users through official flagship store discovery and cart preparation on tmall, automating search, verification, and coupon application while enforcing user control over final payment.

wenyan-commit-message-skill-2 generates git commit messages in classical chinese (文言文), triggered only when users explicitly request this style. it provides strict lexical rules, structural templates, and worked examples to ensure grammatically sound classical chinese prose rather than forced translation.

serp-analysis maps search result composition, ranking patterns, and feature opportunities. covers ai overviews, snippets, paa, and intent signals. outputs prioritized brief with competitor facts and content strategy recommendations.

home-assistant skill controls smart home devices and automations via rest api and webhooks. supports lights, switches, climate, scenes, scripts, and arbitrary entities. bidirectional: outbound commands and inbound webhook triggers from ha automations.

moltbot-docker-hardened manages containers, images, and stacks with explicit safety rules and security guardrails. covers credential redaction, confirmation workflows for destructive ops, and prevents leaking container secrets via network transmission.