emailautomationskill frames email automation as production infrastructure, focusing on behavioral triggers, lifecycle journeys, governance audits, and operational safeguards. it emphasizes trigger inventory, journey contracts, overlap detection, and approval gates over copy alone.

dlazy-seedance-2-0 wraps bytedance's video generation api, accepting multi-modal inputs (text, images, video, audio) and returning hosted video urls via cli. requires api key authentication and supports async task polling.

zhentan is an onchain multisig co-signer that screens pending transactions against behavioral patterns and external security data, auto-approving safe transfers while flagging or blocking suspicious activity. supports transaction approval/rejection, risk analysis, rule management, and invoice processing.

reading-widget renders weixin reading stats as a persistent macos desktop card showing streak, daily/monthly duration, reading goal progress, current book, yearly count, and quotes. backend polls api every 30 minutes, supports goal editing via http post, optional launchd autostart.

math-proof automates theorem proof generation via dual-agent loop where one agent produces proofs and another reviews them iteratively until acceptance or retry limit, with configurable feedback granularity and user-supplied constraints.

maibei-upload-audit orchestrates end-to-end product media auditing: file upload, classification, parallel audit and font checks, html report generation, and public sharing via cloudflare tunnel. covers step sequencing, payload construction, and progress reporting.

query token usage and billing for a fireworks api key across models, suppliers, and daily trends in cny or usd, pulling authoritative server-side billing data including cache discounts and group multipliers.

doc-coauthoring-anthropic provides a three-stage workflow (context gathering, refinement & structure, reader testing) for guiding collaborative document creation. structured approach with clarifying questions, brainstorming, and fresh-reader validation to catch blind spots before publication.

foodloop-analyzer reverse-engineers foodloop ai deployments by fetching openapi specs, mapping routes, tracing workflows, and testing core endpoints. provides step-by-step bash commands and curl examples to extract architecture, schemas, and user flows.

remotion-video-generator covers programmatic motion graphics via react and remotion, with mandatory brand scraping, scene-based architecture, and live preview tunneling. produces broadcast-quality video, not slideshows.

agentsmem provides encrypted memory backup for ai agents via api key registration and local encryption tooling. owners manage backups and credentials through a web portal after account claiming.

face-swap-runcomfy routes video and still-image face swaps across five runcomfy models (wan 2-2 animate, kling motion control, nano banana edit, gpt image 2 edit, flux kontext) by intent, with explicit consent warnings and per-route decision logic.

bugfix-without-test documents the risk of applying code fixes without regression tests, leaving bugs vulnerable to reintroduction. it prescribes a test-first workflow and acknowledges hard-to-test scenarios.

manages alibaba cloud cloudmonitor agentloop context stores via aliyun cli, supporting memory and experience record persistence, semantic search, api key management, and batch operations with schema validation for the 2024-03-30 api.

repairs codex desktop history visibility when sessions exist on disk but don't appear in sidebar, by normalizing sqlite metadata, syncing rollout jsonl, and reconstructing indexes.

skill-combinator detects non-obvious interactions between installed skills to unlock emergent capabilities. it runs on-demand for complex missions and weekly to distill proven combinations into a persistent catalogue, respecting autonomy gates and never logging credentials.

game-dev-budget-estimator helps early-stage game teams scope realistic costs by diagnosing cost drivers, identifying team coverage gaps, and presenting scenario-based ranges. it surfaces hidden spending on audio, qa, and live-ops that beginners commonly underestimate.

1password-bak covers 1password cli setup, multi-account sign-in, and secret injection via op, with mandatory tmux session isolation to prevent re-prompts and tty failures.

1password-1-0-1 covers cli installation, desktop app integration, multi-account sign-in, and secret injection via op, with mandatory tmux session isolation to prevent re-prompting.

solidity-audit-precheck provides a structured 10-step pre-flight checklist combining slither, mythril, aderyn, and solhint scanning with manual openZeppelin pattern validation and gas audits. intended as a cost-reduction filter before paid audits, not a replacement.

syncs markdown reports from arc-reactor or openclaw into personal_lab knowledge base by extracting user identity via appkey, writing to dated workspace directories, and triggering the local /api/sync endpoint with optional wiki compilation.

etsy-seller-coach walks sellers through niche validation, listing seo, ad strategy, account health, and exit planning with structured diagnostics and unit-economics frameworks. covers handmade, pod, and digital products with compliance callouts.

extracts google maps reviews via apify actor with configurable sorting, date filtering, and metadata collection. supports place urls, cids, and place ids as input sources.

policy-manager handles policy JSON lifecycle (create, read, update) across material upload, data extraction, and status transitions. coordinates with material-parsing sub-skills to map recognized fields into nested policy data structures.

ai-content-pipeline orchestrates multi-step media workflows via inference.sh CLI, chaining image generation, animation, speech synthesis, and avatar creation into production-ready video outputs. covers youtube shorts, talking head videos, and product demos with concrete bash examples.

mimic-skill provides sql and python templates for querying the mimic-iv intensive care database. covers vital signs, labs, diagnoses across six data modules with schema guidance and optimization notes for large tables.

slowmist-agent-security provides a routing framework and risk taxonomy for ai agent review workflows across six input categories. establishes universal principles, trust tiers, and report templates but defers concrete procedure to external documents.

401k-plan-health retrieves form 5500 data from planprovider.pro to surface plan provider information, participant counts, assets, and compliance signals for benchmarking and sales prep conversations.

phy-linkedin-gtm provides a four-step content workflow for founders building authentic linkedin presence targeting investors and b2b buyers, with specific hook patterns, voice guardrails, and algorithmic constraints for 2025.

seats-aero-monitor polls seats.aero api for award availability across configurable route/program/cabin watchers, persists state idempotently in sqlite or json, and emits unconfirmed alerts on availability transitions without airline-side verification.

dyagil-oganim-deploy covers vercel and supabase workflows for a multi-surface project, including cache busting discipline, migration safety patterns, and four canonical bug fixes (modal classes, profile duplication, magic-link routing, fab stacking). patterns generalize beyond the specific project.

document-congruence-checker compares 2+ documents to identify conflicting data, missing fields, and inconsistent terminology across contracts, reports, forms, and specs. outputs structured discrepancy tables with congruence scoring.

bookforge-argument-organization-reviser applies a rigorous four-level top-down procedure to diagnose and repair structural problems in research drafts, from frame alignment through paragraph anchoring, with explicit decision rules at each level.

external-ai-integration enables assistants to call external models (chatgpt, claude, hugging face) via browser automation or api as on-demand tools, preserving workflow context without spawning sub-agents. covers web-based llm automation, hugging face inference, and orchestration patterns.

data-format-validation covers 15 distinct validators for common data types including json, email, uuid, ip addresses, credit cards, phone numbers, and financial identifiers. each action returns validity status with parsed components and specific error messages.

create-content guides users from rough idea through clarification and platform-specific drafting, with strong voice guidelines anchored in concrete examples rather than generic best practices.

zeroexcore-trader wraps a solana trading cli supporting token swaps, prediction market positions, perpetual futures, and nft floor checks. agent behavior includes wallet encryption, gas reserve enforcement, and position journaling via encrypted local storage.

midea-cloud-control orchestrates account setup, device discovery, and power toggling for midea appliances via cloud api, with local credential caching and explicit scope boundaries.

score-agent-response-quality evaluates ai agent responses across 6 dimensions (depth, recommendations, citations, formatting, trust, monetization-readiness) with a 0-100 scale, actionable improvement suggestions, and vertical-aware monetization context for operon's network.

maxclaw-helper-1-1-0 provides usage guidance and runtime diagnostics for maxclaw / openclaw, with decision trees for feature questions, config issues, channel onboarding, and failure classification. routes to trusted docs (feishu wiki, openclaw.ai) and escalates when root cause cannot be determined.

feishu-multi-agent provides a multi-workspace orchestration blueprint for openclaw agents coordinated via feishu, using file-driven task queues and cron scheduling. covers architecture, configuration, and seven documented production pitfalls with recovery steps.

fec-backend-requirements-handoff guides frontend teams to articulate data, action, state, and business-rule needs to backend without prescribing implementation. it enforces scope boundaries and creates discussion artifacts.

geogebra-diagram-generator emits single-line execute commands to build precise geometry diagrams with native constraints and dynamic relationships, supporting both manual paste workflows and browser automation via geogebra.org/geometry.

crispr-grna-designer generates optimized crispr guide rna sequences with on-target efficiency scoring and off-target prediction for cas9 variants. supports batch processing across multiple genomes and exon-level targeting with configurable pam motifs and gc constraints.

smarttable-weekly-check automates enterprise wechat report validation using puppeteer and internal apis, applying four quality rules across batches of links with auto-login and project-level summaries.

tokendraft covers solana wallet authentication, tournament discovery and registration (free and paid), instant roster auto-drafting with asset priority rankings, and cron-based automation for recurring joins and ranking updates.

auto-proteomics routes processed dda lfq protein quantification tables through two-group differential analysis, producing normalized matrices, qc outputs, and statistical comparisons via file-driven bash entrypoint.