genkit

Build production-ready AI workflows using Firebase Genkit. Use when creating flows, tool-calling agents, RAG pipelines, multi-agent systems, or deploying AI to Firebase/Cloud Run. Supports TypeScript, Go, and Python with Gemini, OpenAI, Anthropic, Ollama, and Vertex AI plugins.

Install with Tessl CLI

npx tessl i github:supercent-io/skills-template --skill genkit

What are skills?

1.13x

Quality

88%

Does it follow best practices?

Impact

92%

1.13x

Average score across 3 eval scenarios

SKILL.md

Review

Evals

Evaluation results

100%

16%

Product Review Analysis Pipeline

Flow definitions with Zod schemas and null output handling

Criteria

Without context

With context

Zod input schema

100%

Zod output schema

100%

ai.defineFlow usage

100%

Null output guard

100%

No hardcoded API keys

100%

Env var API key

100%

Correct model reference

100%

ESM module type

100%

Correct package names

100%

genkit start with command

100%

inputSchema and outputSchema on flow

100%

Without context: $0.8083 · 3m 39s · 42 turns · 50 in / 10,676 out tokens

With context: $0.8521 · 3m 20s · 33 turns · 282 in / 11,162 out tokens

78%

Customer Support Assistant with Tool Lookup

Tool-calling agent with ai.run() tracing

Criteria

Without context

With context

ai.defineTool name field

100%

ai.defineTool description field

100%

Tool inputSchema

100%

Tool outputSchema

100%

Async tool handlers

100%

Tools array in generate

100%

returnToolRequests false

ai.run() for custom steps

100%

Agent wrapped in defineFlow

100%

Flow Zod schemas

100%

55%

No hardcoded API keys

100%

Null output guard

GENKIT_ENV=dev documentation

60%

Without context: $0.9040 · 3m 43s · 43 turns · 49 in / 11,980 out tokens

With context: $0.8387 · 3m 24s · 29 turns · 34 in / 11,511 out tokens

98%

Long-Form Article Generation Pipeline

Streaming flows and multi-agent pipeline separation

Criteria

Without context

With context

streamSchema defined

100%

sendChunk usage

90%

100%

generateStream used

37%

75%

Separate specialist subflows

100%

Orchestrator delegates to subflows

100%

Dotprompt file created

100%

Dotprompt YAML frontmatter

100%

Dotprompt loaded in code

100%

Flow Zod schemas

100%

No hardcoded API keys

100%

genkit start with command

100%

Null output guard

40%

100%

Without context: $1.5063 · 5m 1s · 49 turns · 102 in / 15,997 out tokens

With context: $1.8703 · 7m 31s · 50 turns · 341 in / 25,489 out tokens

Evaluated: 3 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Product Review Analysis Pipeline Customer Support Assistant with Tool Lookup Long-Form Article Generation Pipeline

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.