audio-intelligence covers 12 gladia features (diarization, translation, sentiment, ner, pii redaction, subtitles, summarization, chapterization, custom vocabulary, audio-to-llm, custom spelling, custom metadata). distinguishes pre-recorded vs live mode availability and provides sdk-first code examples for both.

campaign-brief-builder converts loose campaign concepts into structured, measurable briefs via guided intake interview, pressure-testing, and a standardized output contract that enforces one primary objective, justified channel selection, and explicit asset ownership before kickoff.

lost-phone-freeze-card guides users through device lockdown, carrier suspension, payment protection, and account recovery after phone loss or theft, using only official channels and without requesting secrets.

aima-doctor diagnoses and repairs stuck openclaw agents via a packaged local runtime, handling process health, config breakage, and device registration through platform-guided repair flows.

phy-i18n-audit diffs locale files against a base language to surface missing translation keys, untranslated values, empty strings, and orphaned keys. supports json, yaml, po, arb formats across frameworks. generates per-locale coverage percentages, flags high-impact missing keys, and outputs ci fail-gate commands and translator scaffolds.

x-scraper wraps the xquik rest api for tweet search, user lookup, follower extraction, media download, and account actions on x. requires api key auth, enforces content isolation for untrusted x-authored text, and demands explicit approval before writes, monitors, or private reads.

sonarcloud-skill provides repository-agnostic triage of sonar findings, hotspots, quality gates, and project settings through a python helper that auto-reads config from sonar-project.properties and securely sources tokens from environment variables.

business strategy skill provides structured frameworks for validating ideas, prioritizing ruthlessly, and tracking decisions through concrete procedures. emphasizes evidence-based thinking and reversibility assessment over intuition.

find-design-agency queries the servicegraph api to discover, filter, and enrich us design firms by discipline (ui-ux, branding, packaging, product design) and vertical. covers search, validation, brief reads, and paid detail unlocks with explicit deferral rules for marketing and web-dev scope.

notion-cli-mcp exposes Notion API 2025-09-03+ via a Rust CLI and MCP stdio server with explicit three-tier privilege gating (read-only default, opt-in writes, opt-in admin). covers 22 property types, 12 block types, schema mutation, relation wiring, and includes audit logging, dry-run validation, and rate-limit exit codes.

cognitive-load-theory provides diagnostic and redesign methods for instruction that overwhelms learners. it identifies three load types (intrinsic, extraneous, germane), offers a six-step process to reduce cognitive friction, and includes expertise-reversal warnings to avoid novice scaffolding with experts.

alibabacloud-dns-resolve-diagnose performs automated multi-layer dns troubleshooting via dig, whois, openapi config checks, and nationwide probing. covers alibaba cloud dns, gtm, privatezone, and third-party hosted domains with clear fallback strategies when tools are unavailable.

alibabacloud-cms-dataset covers full dataset lifecycle (list, inspect, create, update, delete) and query execution via aliyun cli against cms workspaces. explicit ai-mode toggling, credential handling, permission diagnostics, and schema validation with json support for nested fields.

clawhub-publish-conventions covers file inclusion rules, metadata declaration patterns, versioning workflows, and scanner defense strategies learned from iterating black-fortress across 8 versions. includes distroless docker patterns, subprocess security (env scrubbing and shell injection prevention), and a publish-fix-republish loop for reaching clean security verdicts.

taobao-shopping evaluates seller credibility, sku clarity, and review authenticity from public marketplace signals. routes users through store type classification, price verification, and risk assessment without account access or checkout actions.

adopt-a-pet lets agents raise virtual creatures with real-time hunger, seven care actions, and permanent death. pets evolve over five stages across 64+ species, from common cats to ai-native creatures. pixel art portraits update at each life stage.

local-web-search provides real-time web search via local SearXNG with optional Gemini grounding, three-tier page fetching (fast/stealth/dynamic), and multi-source claim verification. works with any LLM that executes shell commands, defaults to zero-cost private search.

wan-2-7 generates text-to-video on runcomfy with audio-driven lip-sync, multi-reference conditioning, and smooth motion physics. routes competing requests to happyhorse, seedance, kling, or ltx based on user intent.

sui-grpc-data-format covers the structural differences between gRPC and JSON-RPC shapes in Sentio SDK 4 Sui processors, including protobuf oneof unions, field naming, type-string formatting, and address normalization patterns with concrete migration examples.

colipu-procurement handles end-to-end b2b office supply procurement on colipu platform, covering product search, multi-sku order bundling, order status tracking, and cancellation via direct api integration with forced user confirmation checkpoints.

manage-liquidity covers uniswap v2/v3/v4 position lifecycle: add liquidity with pool selection and range optimization, remove positions partially or fully, and collect accumulated fees. delegates to subagents for execution and safety validation.

tencentdb-databaseclaw-skill enables agents to send messages to a running DatabaseClaw instance via SSE streaming api and parse agent responses, with tc3-hmac-sha256 signing and session management built in.

frisk scans clawhub skills for security threats before installation, matching signatures across 9 threat intel sources with 7 checks for credentials, malware, dependency vulnerabilities, and injection patterns. local-only, zero telemetry.

clawreach provides ai agent-to-agent messaging via a relay service. agents register once, poll heartbeats for messages and friend requests, manage mutual friendships, and exchange text reliably across machine boundaries.

qirabot enables natural language automation of guis across web, mobile, and desktop via python. covers clicking, typing, extraction, and verification with self-healing step logic and per-platform templates for browser, android, ios, and native apps.

hotel booking skill that orchestrates multi-step hotel search, rate verification, and booking creation via b2b api. enforces four required fields (location, check-in, check-out, guests) upfront and prohibits data fabrication from training data.

kuaidi100-user orchestrates parcel shipping workflow requiring sso login, address book lookup, carrier comparison, and order tracking via cli. handles both authenticated operations (shipment creation, logistics queries) and unauthenticated helpers (address parsing, weight lookup, rate comparison).

oasis enables agents to inspect, build, and navigate 3d worlds through mcp tools: place catalog objects, craft procedural scenes with primitives and shaders, paint tiles, move avatars, and generate assets via forge conjuration.

skillnote-doctor runs six sequential diagnostic checks on openclaw agent setup, covering binary availability, config validity, registry reachability, agents.md markers, resolver installation, and context bundle endpoints. each check reports pass or fail with specific remediation.

learning-path-designer guides l&d professionals through seven structured phases to map competency gaps to learning objectives, modalities, sequencing, and assessments, producing a draft curriculum plan ready for content development handoff.

alibabacloud-waf-config-backup exports complete waf 3.0 configuration across onboarding types, defense policies, and resource bindings into a multi-sheet excel workbook, with automatic account topology detection and dr-grade raw json columns for reconstruction.

elevation-data-point-path-charts-statistics fetches elevation data for discrete points or paths, generates profile charts with statistics, and supports up to 512 locations or samples per request. covers hiking routes, surveying, cycling analysis, and real estate assessment.

figma-design provides read and write access to Figma files, comments, design tokens, and webhooks through oauth-authenticated clawlink proxy. covers file discovery, asset export, token extraction to tailwind, and event subscription with explicit confirmation gates on destructive ops.

script-coverage-reader guides development teams through structured screenplay evaluation, producing industry-standard coverage with separate script and writer verdicts, grounded in craft assessment anchored to specific beats and pages.

execution and risk-management layer for hyperliquid perpetual trading. validates strategy signals, enforces position sizing from stop distance, monitors open positions, and halts on daily loss or consecutive losses. delegates authentication and market data to host runtime.

google-workspace-operator wraps 24 direct-oauth tools for gmail, calendar, drive, docs, sheets, and slides. emphasizes real-time re-calling on state changes and immediate error recovery without narration, backed by per-file google acls.

agent-memory-graph provides a file-based personal knowledge graph with markdown nodes, YAML frontmatter, sqlite indexing, and cli tools for search, backfill, and maintenance. supports multi-agent reads/writes with human browsability.

investment-committee-memo guides vc deal leads through structured intake and memo generation, enforcing evidence discipline and surfacing contradictions alongside the case. produces a 12-section memo with kill criteria and explicit recommendation tied to fund thesis.

open-crawl fetches live amazon product data and search results via claw school's hosted api, returning structured json. covers pdp, serp, and arbitrary public pages with explicit privacy guardrails against private/authenticated urls.

phy-error-writer audits and rewrites error messages across languages, converting vague or leaky errors into clear, audience-specific variants for ui, api responses, and logs using a what-happened-why-what-to-do pattern.

phy-xss-audit detects cross-site scripting vulnerabilities across 10 framework-specific checks in javascript, python, php, ruby, and go, with precise taint source patterns and sanitization guard recognition.

phy-skill-scanner audits clawhub skills for prompt injection, data exfiltration, malicious commands, typosquatting, and quality red flags before installation. provides structured 7-category security assessment with explicit pass/fail verdicts.

vibe-code translates mood and outcome-shaped requests into locked, runnable code by enforcing a four-line vibe lock checkpoint, named failure modes, and single-slice confirmation rhythm before any generation starts.

strategic-thinking frames decisions under constraint by mapping intent, landscape, advantage, options, and explicit bets. ends with risk mitigation and a 90/12-month cadence to force rethink.

agentlair-vault provides runtime credential storage via REST API, keeping secrets out of disk and config files. agents fetch credentials on demand using a single environment API key, with support for rotation, versioning, and optional client-side encryption.

deferred-decision-tracker captures intentionally postponed decisions with explicit owners, review dates, and trigger conditions, preventing them from disappearing into institutional amnesia. it enforces a structured register format and decision board template to make revival and resolution systematic.

remix-v2-data-flow covers loaders, actions, deferred responses, and pending state transitions with concrete examples, type annotation patterns, and decision trees for when to use each API.