Skills
Install into Cursor, Claude Code, or Codex with one command.
4 free · 17 premium
Plan AI SDK Agent
Inspects your repo and writes a Markdown blueprint covering architecture, tools, state, security, tests, and build order.
AISDK Agents
Architecture guidance, installed-version API checks, and starter shapes distilled from 125 AI SDK patterns.
Build AI SDK Agent Evals
Adds datasets, deterministic tests, model-graded scorers, and local or CI commands around an existing agent.
Observe AI SDK Agent
Wires production visibility into your agent: native telemetry extended with tool spans, per-user cost attribution, privacy-safe capture, and failure traces that become eval fixtures.
Optimize AI SDK Agent
Baselines the agent, tests focused changes, reruns checks, and keeps improvements supported by the results.
Write Agent Prompts
Rewrites the prompts your agent actually runs on: a repo-wide audit of contradictions and token waste, prompts as versioned code, and before and after evals proving every change.
Better Chat UI
Opinionated principles for AI SDK useChat surfaces: parts, statuses, streaming markdown, scroll, composer, and persistence.
Better Generative UI
Maps tool calls, structured objects, and data parts onto typed components — skeletons, approvals, artifacts, no raw JSON.
Ingest Messy Documents
Gets messy sources into faithful structured text: measured text-layer quality, OCR and vision strategy per document class, lossless tables, provenance for citations, and incremental re-ingestion.
Ground AI SDK Agent
Adds retrieval your agent can defend: corpus-derived chunking, hybrid search and reranking, citations in the UI, and retrieval-quality evals.
Agent Memory
Gives your agent durable per-user memory: an evidence-based write path, budgeted recall injection, decay, user-visible deletion, and recall-quality evals.
Build MCP Server
Turns your API into an MCP server models use well: task-shaped tools, pagination and truncation discipline, real auth, and an eval harness measuring task success.
Extract Structured Data
Turns documents and text into typed records you can defend: schemas designed for the model, bounded validation and repair, confidence routing to review, and per-field accuracy on a golden set.
Ship AI SDK on Cloudflare
Deploys your agent to Cloudflare Workers with streaming that survives the edge: runtime architecture from your agent's shape, Durable Object state, AI Gateway, and a deployed smoke test.
Ship AI SDK on AWS
Deploys your agent to AWS in your own IaC dialect: streaming that survives every layer, durable workflows with checkpoints, and Bedrock or external providers wired with IAM, verified against the deployed URL.
AI SDK on Mobile
Puts streaming AI chat in your React Native or Expo app: keys stay server side, the transport matches your installed versions, and backgrounding and flaky networks are designed states.
Migrate to AI SDK
Moves your codebase to AI SDK primitives without a rewrite: full call-site inventory, a safety net captured before any change, and parity proven against goldens before the old stack is removed.
Setup
Grounds your coding agent in a repo in one command: explores the codebase, runs one interview over every decision, then dispatches to the bootstrap and design engines.
Bootstrap AGENTS.md
Emits AGENTS.md, CLAUDE.md, CONTEXT.md, docs/agents, plans, and portable plan-mode and code-review skills, all rewritten from evidence gathered in your repo.
Design MD
Creates or updates a living DESIGN.md design-language contract plus a portable frontend-design skill. Documents what exists or proposes a direction, never restyles.
Frontier Interview
Runs multi-decision interviews as a frontier walk over a decision tree: each round asks every settled question with a recommended answer, then recomputes.
Get every skill
Premium unlocks all 17 paid skills and the 125-pattern source library.