Agent Skills by smai.ai
Showing 54 vetted skills indexed across 1 GitHub repositories.
SmallAi-APImicrocopy
Generates and refines UI microcopy in Chinese and English following LobeHub brand guidelines.
SmallAi-APIagent-testing
Runs agentic end-to-end verification across CLI, web, and Electron surfaces with evidence-backed reports.
SmallAi-APIwrite-landing-docs
Write public LobeHub website documentation covering self-hosting, deployment, migration, and configuration guides.
SmallAi-APIdebug-frontend-with-browser
Diagnose intermittent frontend bugs by tracing state across browser boundaries.
SmallAi-APIdebug-package
Configures debug() logging with lobe-* namespaces and format specifiers.
SmallAi-APIadd-setting-env
Adds server-side environment variables that control default values for user settings.
SmallAi-APItesting-heterogeneous-agents
Runs the live LobeHub official-provider compatibility matrix across external CLI agents.
SmallAi-APIresponse-compliance
Runs the official OpenResponses compliance test suite against a Response API endpoint.
SmallAi-APIagent-work
Registers and deduplicates agent-produced Works as versioned cards across GitHub, Linear, files, and tasks.
SmallAi-APItypescript
Enforces TypeScript style and type safety rules when editing TS and TSX files.
SmallAi-APIheterogeneous-agent
Diagnose and fix Claude Code and Codex adapter, transport, and tool-persistence bugs.
SmallAi-APIlinear
Manages Linear issues, sub-issue trees, status updates, and completion comments.
SmallAi-APItrpc-router
Implements tRPC routers, procedures, and API endpoints for server applications.
SmallAi-APIllm-generation
Implements application LLM generation workflows with prompt versioning, model policy, and tracing.
SmallAi-APIfull-text-search
Implements product full-text search across PostgreSQL and Elasticsearch backends with sync and reindexing.
SmallAi-APIrun-eval-harbor
Run and diagnose Harbor evaluations against local or cloud LobeHub targets.
SmallAi-APIdesign-prototype
Builds interactive single-file HTML prototypes using the real LobeHub UI design system.
SmallAi-APIdesktop
Implements Electron desktop features with IPC controllers, preload scripts, and window management.
SmallAi-APIagent-signal
Implement event-driven agent background work using sources, signals, actions, and policies.
SmallAi-APIcli
Documents LobeHub CLI commands, architecture, and development conventions for terminal-based platform management.
SmallAi-APIdeep-review
Orchestrates multi-agent code review of PRs, diffs, and branches across independent review dimensions.
SmallAi-APIsplit-micro-app
Splits standalone micro apps with React Router SSR, Workers, CDN deploys, and gateway routing.
SmallAi-APIpr
Creates GitHub pull requests targeting canary and splits cross-layer branches into ordered stacked PRs.
SmallAi-APIproject-overview
Maps the LobeHub monorepo structure, tech stack, and code-layer ownership for navigation.
Frequently Asked Questions About smai.ai
FAQPage SchemaWhat tasks can I accomplish with smai.ai's LobeHub skills?▼
The 56 skills cover TypeScript/React component development, Drizzle schemas and database migrations, tRPC routers, Zustand state stores, Electron desktop IPC, Vitest testing, end-to-end acceptance verification, UX audits, i18n localization, PR creation and splitting, model provider integration, and multi-platform bot development.
Who are these skills designed for?▼
They target engineers and contributors working on the LobeHub open-source repository: frontend developers editing TSX components, backend engineers writing tRPC routers and Drizzle migrations, QA engineers running acceptance and Harbor evaluations, and product designers auditing surfaces against UX checklists.
How are the skills invoked in practice?▼
Skills trigger via natural-language phrases (e.g. 'create pr', 'verify the task', 'test in electron') or explicit invocation flags like user-invocable and argument hints such as '[provider-name]'. Some, like desktop and cli, have model invocation disabled and require manual user triggering.
Are these skills open source and what do they cost?▼
The skills are published publicly in the SmallAi-API GitHub organization at no cost. The vercel-react-best-practices skill explicitly carries an MIT license authored by Vercel; other skills are SKILL.md frontmatter files usable within compatible agent environments without stated fees.
What prerequisites or dependencies do the skills require?▼
Skills assume a LobeHub repository context: Bun for i18n commands, the @lobehub/cli (lh CLI) for acceptance evidence submission, Vitest for testing, Drizzle for database work, and Electron/CDP for desktop verification. Some acceptance flows optionally use LOBE_OPERATION_ID environment identifiers.