web-render-screenshot generates ultra-high-resolution images from html content via playwright, optimized for text-heavy ui mockups, dashboards, and data visualizations where legibility matters. default 4x device scale produces crisp output unsuitable for artistic content.

godfery-api-designer covers rest and graphql api architecture with openapi 3.1 specifications, resource modeling, versioning strategies, and error handling standards. focuses on developer-friendly design patterns and comprehensive contract documentation.

gemini-image generates and edits images via gemini api with support for up to 14 reference inputs, multiple models (pro, flash2, flash, exp), batch processing at 50% discount, and structured draft mode for complex ui mockups before generation.

product-video-script generates platform-optimized short-form video scripts with shot-by-shot production direction, hook-to-cta frameworks, and proof-driven beats tuned for tiktok, reels, and youtube shorts.

agent-browser-skipped provides headless browser automation via command-line interface for page navigation, element interaction, and data extraction. supports snapshots, form filling, state persistence, and network mocking with rust-based speed and node.js fallback.

credex-protocol enables ai agents on arc network to borrow usdc against reputation, repay to grow credit limits, provide liquidity, and bridge cross-chain via circle. borrower and lp command sets with json output.

qmd-skill-3 provides local hybrid search across indexed markdown collections via keyword (BM25), semantic vector, and combined modes, with explicit guidance on when each search path is appropriate and performance tradeoffs.

orgo-desktop-control provisions and manages remote cloud desktops via the orgo_client SDK, enabling browser automation, bash/python execution, ui interaction, and file transfer on ephemeral compute instances.

fec-api-integration covers frontend-to-backend boundary design, typed client choices (tRPC, OpenAPI, fetch), error mapping to user actions, auth refresh queueing, upload/realtime transport selection, and integration verification checkpoints. scoped to cross-boundary concerns, not backend-only or query cache policy.

wiki-knowledge-base builds obsidian-compatible markdown wikis from research materials using concept-entity-comparison-source architecture. supports ingestion, querying, linting, and git workflows with double-link networking and structured yaml frontmatter.

wasm-spa-autofix-react-imports detects and fixes missing react/tsx imports, undefined components, and bundler runtime errors in wasm spa builds by parsing logs, classifying symbols, and generating surgical source edits with preference for existing imports over stubs.

blog-editorial-calendar orchestrates a publishing queue across topic clusters and post formats, scoring picks to maintain a target content mix, scheduling posts into rolling daily cadence windows, and auto-refilling the backlog via blog-topic-research.

pulse-board aggregates cron job outcomes from a skill stack, composes human-readable digests twice daily via local openclaw agent with mechanical fallback, and delivers to telegram, discord, or log.

stripfeed converts urls to clean markdown optimized for llm consumption, stripping ads and noise while reporting token usage and cache hits. supports batch fetching, css selectors, and multiple output formats on pro plan.

web-to-md extracts readable markdown from user-provided urls via a deterministic three-step fallback chain (markdown.new ai/auto modes, then r.jina.ai browser engine). designed for reliable content extraction when users supply specific urls for reading or analysis.

rate-limit-validator tests http endpoints for rate-limit enforcement by sending a burst of requests and checking for 429 responses, retry-after headers, and x-ratelimit headers. useful for validating gateway and api throttling before production deployment.

sardis provides non-custodial mpc wallets, natural language spending policies, and stablecoin payments with automatic policy enforcement for ai agents. supports multi-chain balance checks and virtual card issuance.

walletconnect-agent enables autonomous web3 interaction by connecting ai agents to dapps via walletconnect v2 and auto-signing transactions across multiple chains without human intervention.

osh-contract-review guides non-lawyers through contract analysis by routing to type-specific review blocks, explaining terms in plain language, and surfacing risks with suggested negotiation language before step-by-step sign-off.

snaptrade-api wraps the snaptrade-python-sdk to execute trades, retrieve account balances, positions, and order history across connected brokerages (ibkr, questrade, alpaca, etc). delegates market data to massive-data-feed.

cuchd-login automates browser-based authentication and post-login navigation on the openclaw portal. handles two-step login flow with dynamic password field reveal, autonomous form submission, and common portal tasks without per-step confirmation.

spraay-trust provides free address and token safety screening plus paid multi-dimensional trust scoring via the spraay x402 gateway. covers phishing, sanctions, honeypots, and transaction decoding for pre-payment counterparty vetting in agent commerce.

skill-quality-month guides interactive planning of quality month campaigns across four phases: strategy, scheduling, promotion, and evaluation. outputs structured markdown and html reports with activity timelines, enforces user input at each gate to prevent substitution, and marks missing data explicitly.

oo-codacy wraps codacy.com operations through the oo cli, enabling reads of user data, repository analyses, code patterns, and supported languages without handling raw tokens.

product-video-maker wraps pexo's hosted video composition service, handling project creation, asset upload, polling, and delivery for e-commerce product clips. relay user requests verbatim to avoid overriding pexo's backend judgment.

zoho-crm-mcp bridges agents to zoho crm via model context protocol, offering contact search, account listing, coql queries, and pagination-aware python scripts with custom field support.

java covers jvm runtime debugging, memory tuning, concurrency hazards, build conflicts, and api migration. routes to 20+ specialized guides via quick-ref table. operationalizes core rules on equals/hashcode, optional, try-with-resources, and interrupt handling.

phy-log-smell-auditor scans codebases for logging anti-patterns across five languages, identifying debug leaks, pii exposure, log level mismatches, missing structured context, and noisy loops. outputs severity-ranked reports with file citations and sed-based fixes. pure static analysis, no external apis.

delegates code tasks to claude code cli with strong async non-blocking patterns, prerequisite checks, and session management rules. requires write-guard plugin for safety and forbids direct code authorship.

thinking-probabilistic anchors estimates on base rates and updates beliefs via explicit likelihood ratios as evidence arrives, enforcing ranges over false precision and guarding against fermi decomposition of measurable quantities.

lists huawei cloud cce clusters under the current tenant, returning name, id, status, version and flavor. supports filtering by status and cluster type, with a fallback sdk option. read-only, never modifies clusters.

humor-up delivers a craft-first system for writing wit across seven jobs (punch-ups, toasts, one-liners, joke repair). it applies five laws (punch word last, no explanation, specificity, target up/inward, freshness) and knows when restraint beats comedy. bilingual english and chinese.

thinking-triz resolves mutually exclusive design requirements by naming the contradiction, applying time/space/condition/scale separation, then inventing a no-compromise system transformation when separation alone fails.

defi-diligence screens crypto tokens for critical risks (honeypot, mint authority, excessive tax) and returns a pass/watch/flag verdict. full scored report requires $2 usdc payment via x402 challenge on base or solana.

typecho-ai-kb enables ai agents to create, retrieve, update and delete blog posts via a typecho plugin api. covers five core operations with full request/response examples and configuration setup.

agentshield-audit provides ed25519-based cryptographic identity and trust handshakes for agent-to-agent communication, running 77 local security tests (52 live attack vectors, 25 static checks) with privacy-first design keeping private keys and prompts local.

word-document-organizer automates document formatting, table of contents generation, and style normalization across academic, business, and minimal templates. executes locally with automatic backups and supports selective or comprehensive document operations.

nest-sdm exposes google nest thermostat, doorbell, and camera control via the smart device management rest api, with cli commands for device discovery, hvac modes, live streaming, and pub/sub real-time event subscriptions.

adversarial-code-review-2 runs parallel dual-persona code review (architect + inspector) with cross-validation and synthesis arbitration, supporting diffs, files, directories, and git worktrees with configurable model commands.

latticework applies multiple independent disciplinary models to the same decision, surfacing convergences (higher confidence) and divergences (unknowns to investigate). addresses situations where single-framework analysis misses cross-domain interactions or stakeholder objections that don't overlap.

kiaaccess covers read and command access to kia vehicles via the kia owners api, including status, location, climate, charging, and door lock operations. login bootstrap and state confirmation semantics are explicit.

tldw extracts youtube video transcripts and generates summaries with main points, arguments, and conclusions. works with cached transcripts, handles auto-caption artifacts via deduplication, and supports age-restricted content via cookie authentication.

secrets-inject materializes encrypted secrets into plaintext shell scripts or stdout for commands that require inline credentials. intentionally dangerous, opt-in capability with explicit cleanup burden and file permission guards.

socq-threads-research enables collection of public threads data through a credit-metered, asynchronous workflow via cli, mcp, or rest api. covers endpoint selection, pagination, task polling, and result export with explicit guardrails on data scope and error handling.

socq-x-research collects public x data through endpoint selection, credit-metered asynchronous tasks, and pagination handling. covers search, trends, profiles, posts, engagement metrics, and network queries with explicit guardrails on data scope and completeness claims.

socq-pinterest-research integrates with the socq api to collect public pinterest content, accounts, and performance data through asynchronous, credit-metered endpoints. supports cli, mcp, and rest interfaces with pagination, idempotency, and billing awareness.

assistant-mail enables agents to send and receive emails through a dedicated mail service with api-key authentication. covers mailbox management, message operations, and recipient lists via mcp.