Data framework for building LLM applications with RAG. Specializes in document ingestion (300+ connectors), indexing, and querying. Features vector indices, query engines, agents, and multi-modal support. Use for document Q&A, chatbots, knowledge retrieval, or building RAG pipelines. Best for data-centric LLM…
Wires up an Airbyte connector for use in a PydanticAI or Claude SDK agent, or any other framework via agenttool. Generates auth config, connector initialization, and toolutils- or agenttool-decorated tool functions. Use when adding a connector to an agent or setting up a new agent with a connector.
Builds a complete agent with multiple Airbyte connectors using PydanticAI or Claude SDK. Scaffolds project structure, wires up connectors, composes tools, and creates a run loop. Use when building an agent with multiple connectors or scaffolding a new agent project.
Discovers available Airbyte connectors, explores their entities/actions/schemas, and checks auth requirements. Use when asking what connectors exist, what a connector can do, or what fields/entities are available.
Run APR (Automated Plan Reviser) workflow for FCP specification refinement. Use when asked to run APR, revision rounds, spec review, integrate GPT feedback, or harmonize docs. Handles Oracle remote setup, running rounds, integrating feedback, and syncing README/implementation docs.
Show which Intent Outreach data connectors are configured (have a bring-your-own key) versus skipped, with each one's tier. Use when the user asks which connectors or data providers are set up, whether a specific provider (apollo, hunter, …) is configured, or wants a preflight check before running a campaign.…
Produce a clean export/delivery spec - right formats, sizes, densities, naming, color profile, optimization - for a set of assets. Use when delivering assets to a client or developer.
Critique a design against craft principles - hierarchy, layout, type, color, spacing, accessibility, brand fit. Use when reviewing a draft, a generated image, or a layout before delivery.
Define or extract a design token system (color, type, spacing, radius, shadow) and output it as JSON/CSS variables. Use when setting up a design system or converting a Figma/brand palette to tokens.
This skill should be used when the user asks to "build our strategy map," "draft objectives for the scorecard," "what's our causal chain," or has candidate objectives that need an explicit causal mechanism linking each one to the perspective above it — the actual intellectual core of a balanced scorecard, not just…
This skill should be used when the user asks to "cascade the scorecard to business units," "build a team-level scorecard," or needs corporate strategy-map objectives translated into business-unit or team scorecards via a genuine causal-mechanism test, not just contribution-counting.
This skill should be used when the user asks to "define our perspectives," "set up our scorecard structure," "should we use the standard four perspectives," or needs to decide the perspective set and causal-chain order before any objectives get drafted.