nonprofit-toolkit

Build Slack bots with Bolt for JavaScript on Bun. Use this skill whenever the user wants to add a Slack command, event handler, assistant, modal, interactive component, or any Slack integration. Also triggers on "add a slash command", "Slack notification", "Slack bot", "app_mention", "Assistant API", "Slack event", "weekly digest", "scheduled message", or "reaction handler". Covers both real-time handlers (commands, events, modals) and scheduled/proactive messages.

Answer questions about the AI SDK and help build AI-powered features. Use when developers: (1) Ask about AI SDK functions like generateText, streamText, ToolLoopAgent, embed, or tools, (2) Want to build AI agents, chatbots, RAG systems, or text generation features, (3) Have questions about AI providers (OpenAI, Anthropic, Google, etc.), streaming, tool calling, structured output, or embeddings, (4) Use React hooks like useChat or useCompletion. Triggers on: "AI SDK", "Vercel AI SDK", "generateText", "streamText", "add AI to my app", "build an agent", "tool calling", "structured output", "useChat".

Run compliance-discovery sources against the configured nonprofit entity and report findings. Use this skill when the user asks "are we compliant?", "check our IRS status", "verify our tax-exempt status", "run a compliance check", "is our 501(c)(3) still active?", or when they want to refresh the compliance picture before a board meeting, audit, or grant application. Phase 3 covers IRS TEOS/BMF, public CA AG Registry Search Tool/detail pages, public CA SOS bizfile checks, public CA FTB Entity Status Letter checks, public CDTFA permit verification, and user-assisted authenticated CA portal checks.

Capture the nonprofit's identifying information for compliance discovery and persist it to Secret Manager (entity IDs) and BigQuery (entity attributes). Use this skill when the user is setting up the compliance toolkit for a new nonprofit, when they say "onboard us", "set up compliance", "register our nonprofit details", "configure the compliance module", or when running `compliance-discover` reports that no entity is configured. Also use when the user wants to update the legal name, EIN, fiscal year end, or mailing address on file.

Summarize the configured nonprofit's stored compliance state without running network discovery. Use this skill when the user asks for current compliance status, the latest stored findings, or whether prior discovery found anything that needs attention.

Bulk data enrichment. Adds web-sourced fields (CEO names, funding, contact info) to lists of companies, people, or products. Use for enriching CSV files or inline data. Supports multi-turn: pass --previous-interaction-id from a prior research or enrichment to carry context forward.

ONLY use when user explicitly says 'deep research', 'exhaustive', 'comprehensive report', or 'thorough investigation'. Slower and more expensive than parallel-web-search. For normal research/lookup requests, use parallel-web-search instead. Supports multi-turn: pass --previous-interaction-id from a prior research or enrichment to continue with context.

URL content extraction. Use for fetching any URL - webpages, articles, PDFs, JavaScript-heavy sites. Token-efficient: runs in forked context. Prefer over built-in WebFetch.

DEFAULT for all research and web queries. Use for any lookup, research, investigation, or question needing current info. Fast and cost-effective. Only use parallel-deep-research if user explicitly requests 'deep' or 'exhaustive' research.

Get completed research task result by run ID

Set up the Parallel plugin (install CLI)

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

Check running research task status by run ID

Provision GCP infrastructure for the Donations ETL pipeline. Creates .env from template if needed, guides through API key setup for Mercury, PayPal, Givebutter, and Check Deposits (Google Sheets), then runs the provisioning script.

Create and deploy MCP servers with tools, prompts, and OAuth authentication. Use when building, deploying, or debugging an MCP server, adding OAuth to one, or connecting an MCP client (Claude.ai, Claude Code, ChatGPT) to a custom backend.

Build AI-powered natural language analytics using the Vercel AI SDK ToolLoopAgent pattern with BigQuery. Use this skill whenever the user wants to add a natural language query interface, NL2SQL, data Q&A bot, AI-powered reporting, or any feature where users ask questions about data in plain English and get answers from a database. Triggers on "ask questions about data", "natural language SQL", "AI query", "data assistant", "analytics bot", "chat with data", "query bot for a different table", "add a data source", "bad SQL", "hallucinated columns".

Bootstrap donations-etl on a fresh machine. Use this skill when the user says "bootstrap", "set up from scratch", "fresh install", "new machine setup", "onboard me", "getting started", "install everything", "clone and set up", or asks how to get the project running on a new machine. Also use when the user encounters missing dependencies or environment issues that suggest the project hasn't been fully set up. Covers installing Bun, dependencies, and walks through full configuration and optional deployment.

Create a new data source connector for donations-etl. Use this skill when the user wants to add a new payment platform, bank, fundraising service, or any other donation data source. Triggers on "add connector", "new connector", "add data source", "integrate with X", "connect to X", "import from X", "add X as a source", or when the user mentions a payment platform (Stripe, Square, GoFundMe, etc.) and wants to pull donation data from it. Also use when modifying or extending an existing connector.

Run the ETL job locally with an interactive UI. Use when the user wants to "run ETL", "test ETL locally", "execute donations sync", or "load data from sources". Provides guided selection of mode, sources, and pipeline stages.

Use when implementing any feature, fixing any bug, or investigating any problem — before writing production code. This skill MUST be used whenever: writing new code, debugging issues, a user reports a problem, you see errors in logs, data looks wrong in BigQuery or any destination, a test is failing, behavior doesn't match expectations, or you're about to change existing code for any reason. The core rule is simple: reproduce the problem or define the behavior with a failing test BEFORE writing or changing production code. Use this even for "quick fixes" — especially for quick fixes.

Use when creating or updating command-line utilities, CLI tools, or scripts that parse command-line arguments. Guides through the standard pattern using commander and Zod for type-safe CLI utilities.

Use when deploying the ETL application to GCP Cloud Run. Triggers on "deploy", "push to production", "release to cloud", "update cloud run", or "ship the ETL". Handles Docker image builds via Cloud Build and Cloud Run Job updates.

Query and analyze the donations BigQuery table using natural language. Use this skill whenever the user asks questions about donations, donors, revenue, fundraising metrics, payment sources, transaction history, or wants to explore donation data. Triggers on questions like "how much did we raise", "who are our top donors", "show me donations by source", "what's our monthly revenue trend", "any failed transactions", or any freeform analytical question about the donations dataset. Also use when the user says "query donations", "donations analytics", "run a query", or asks about donor retention, average gift size, or campaign performance.

Generate a formal donation confirmation letter for a donor. Use this skill whenever the user asks to write a letter, confirmation, receipt, acknowledgment, or thank-you for a donor. Triggers on phrases like "write a letter for donor", "donation confirmation letter", "donor acknowledgment", "generate a receipt for donations", "make a letter like we did before", or when someone provides a donor email and asks for documentation of their contributions. Also use when the user forwards a donor's request for confirmation of their giving history.

Use when working with external libraries, APIs, GitHub Actions, or SDKs. Reminds to search for current documentation instead of relying on training data which may be outdated.