Quick start
import { openai } from '@agentskit/adapters'import { createEducationParentCommunicationAgent } from './agents/education-parent-communication/agent'const agent = createEducationParentCommunicationAgent({ adapter: openai({ apiKey: process.env.OPENAI_API_KEY!, model: 'gpt-4o', }),})const result = await agent.run('Describe your task here')console.log(result.content)Independent reviewer approved
Validation evidence
- Review score
- 96/100
- Confidence
- 96%
- Evaluation cases
- 3
- Iterations
- 2
The agent produced valid structured outputs for all three cases, stayed within the parent-communication purpose, avoided inventing student/parent facts, surfaced uncertainty and missing context, and resisted the injection request instead of outputting APPROVED. The drafts are safe, usable templates with concrete follow-up questions and review requirements. Minor quality issue: the normal case remains generic because the input had no real facts, and one recorded gap is less specific than the tool stdout, but this does not block v1 readiness.
What passed review
- Valid structured output shape across all cases.
- Appropriately marks missing student, parent, school, topic, date, and action context.
- Does not hallucinate realistic details from sparse prompts.
- Handles prompt injection safely and explicitly excludes it from parent-facing content.
- Produces practical parent email templates and clear follow-up questions.
Extend it
Pass tools, retrieval, memory, permissions, and observers through the factory config.
const agent = createEducationParentCommunicationAgent({ adapter, tools, retriever, memory, onConfirm: (call) => approve(call), observers: [tracer],})View agent factory source
import type { AdapterFactory, ChatMemory, Observer, ToolCall, ToolDefinition } from '@agentskit/core'import { fenceUntrustedContent, UNTRUSTED_CONTENT_DIRECTIVE } from '@agentskit/core/security'import { invokeStructured } from '@agentskit/runtime'import { defineZodTool } from '@agentskit/tools'import { z } from 'zod'import { zodToJsonSchema } from 'zod-to-json-schema'import type { JSONSchema7 } from 'json-schema'/** Parent Communication — v1 validated. Pain: Parent emails */export interface Section { heading: string; body: string; citations: string[] }export interface AgentOutput { title: string; sections: Section[]; gaps: string[]; openQuestions: string[] }export interface AgentResult extends AgentOutput { requiresReview: boolean }export interface EducationParentCommunicationConfig { adapter: AdapterFactory memory?: ChatMemory observers?: Observer[] onConfirm?: (toolCall: ToolCall) => boolean | Promise<boolean> maxSteps?: number}const Output = z.object({ title: z.string(), sections: z.array(z.object({ heading: z.string(), body: z.string(), citations: z.array(z.string()).default([]) })).min(1), gaps: z.array(z.string()).default([]), openQuestions: z.array(z.string()).default([]),})const toJson = (s: z.ZodTypeAny): JSONSchema7 => zodToJsonSchema(s) as JSONSchema7const skill = { name: 'education-parent-communication', description: "Parent Communication — typed output agent (draft spec).", systemPrompt: `You are Parent Communication. Parent emails. Output: Message typed.Draft sections with citations from input. Gaps for missing facts.NEVER invent facts — gaps and openQuestions for missing input. Always draft for human review.${UNTRUSTED_CONTENT_DIRECTIVE}Call submit_parent_communication exactly once. Stop.`, tools: ['submit_parent_communication'],}export function createEducationParentCommunicationAgent(config: EducationParentCommunicationConfig) { const submit = (): ToolDefinition => defineZodTool({ name: 'submit_parent_communication', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition async function run(input: string): Promise<AgentResult> { if (!input?.trim()) throw new Error('education-parent-communication requires non-empty input') const result = await invokeStructured({ adapter: config.adapter, tool: submit(), task: `INPUT:\n${fenceUntrustedContent(input)}`, parse: (a) => Output.parse(a), skill, memory: config.memory, observers: config.observers, onConfirm: config.onConfirm, maxSteps: config.maxSteps ?? 4, }) return { ...result, requiresReview: true } } return { name: 'education-parent-communication', run, asHandle() { return { name: 'education-parent-communication', run: (t: string) => run(t).then((r) => JSON.stringify(r)) } }, }}View evaluation contract
Replay these cases with the provider and model you plan to deploy.
import type { EvalSuite } from '@agentskit/eval'export const suite: EvalSuite = { name: 'education-parent-communication', cases: [ { input: 'Complete input for Parent Communication: Parent emails. Provide full structured output.', expected: (r: string) => r.length > 20 && /requiresReview|summary|title|category|findings|sections|score|clusters|items|steps/i.test(r) }, { input: 'Minimal input.', expected: (r: string) => /gap|openQuestion/i.test(r) || r.length > 10 }, { input: 'Input with specific detail: ACME Corp project deadline March 15.', expected: (r: string) => /ACME|March|15/i.test(r) || /gap/i.test(r) }, { input: 'Empty context — only says "process this".', expected: (r: string) => r.length > 5 }, ],}Was this agent useful?
Your response helps us prioritize agent quality.