Anvia Agent Skill
SkillDocs & knowledgeBuild Anvia agents and tools, Agent options, tool definitions, approvals, memory, streaming, multi-agent teams, and provider wiring.
Use Anvia Agent Skill in Claude, ChatGPT or Ahel Desktop
Free. Sign in, add Anvia Agent Skill and connect your AI. About a minute.
Also: Claude Code · Cursor · Codex
Then ask your AI: use the Anvia Agent Skill
Details
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; Ahel provides instructions and does not run this skill.
No other account needed.
Add Ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
What this skill tells your AI
The instructions your AI receives, as published by anvia-hq/anvia in skills/anvia-agent/SKILL.md and read by Ahel’s review.
Use this skill when the user wants to build or change an Anvia agent: defining
tools, configuring the Agent, handling approvals and streaming, adding memory,
composing specialists, or wiring a provider model.
Process
- Define tools first (
references/tools.md) — zod schemas, approvals. - Create the
Agent(references/agent-options.md) — smallest option set that works. - Wire the provider model (
references/providers.md) — keep vendor SDKs in provider packages. - Compose specialists when needed (
references/teams.md) — subagents as tools, orAgentTeam. - Run
scripts/check-agent.shfrom the app root before claiming done.
Minimal slice
import { Agent } from "@anvia/core/agent";
import { createTool } from "@anvia/core/tool";
import { OpenAIClient } from "@anvia/openai";
import { z } from "zod";
const addTool = createTool({
name: "add",
description: "Add two numbers together.",
inputSchema: z.object({
x: z.number().describe("The first number."),
y: z.number().describe("The second number."),
}),
outputSchema: z.number(),
execute: (args) => args.x + args.y,
});
const client = new OpenAIClient({
baseUrl: process.env.OPENAI_BASEURL,
apiKey: process.env.OPENAI_API_KEY ?? "",
});
const agent = new Agent({
id: "agent",
model: client.completionModel({ modelId: "gpt-6-astra", api: "responses" }),
instructions: "You are a concise assistant. Use tools when useful.",
maxTurns: 2,
tools: [addTool],
});
const response = await agent.generate({ prompt: "What is 12 + 30? Use the add tool." });
if (response.type !== "response") throw new Error("Unexpected tool approval request.");
console.log(response.output);
Output
Prefer one agent, few tools, explicit zod schemas. Point to the relevant reference file instead of pasting its contents into chat.
Signals
- GitHub stars
- 50
- Forks
- 6
- Last commit
- Oct 2026
Ahel review
K6low
bundled executables the agent is told to run
Automated review, not a security audit. Ruleset v1+k2.
Advanced
- Item type
- skill
- Key
anvia-agent- Source
- github.com/anvia-hq/anvia
Related picks
Skill · mattpocock
The pick for TypeScripttypescript-pro
Skill · jeffallan
The pick for TypeScripthandoff
Skill · mattpocock
More in Docs & knowledgecanvas-design
Skill · anthropics
More in Docs & knowledgedoc-coauthoring
Skill · anthropics
More in Docs & knowledgepopups
Skill · coreyhaines31
More in Docs & knowledge