Quick start
import { openai } from '@agentskit/adapters'import { createOpsAssetInventoryReconcilerAgent } from './agents/ops-asset-inventory-reconciler/agent'const agent = createOpsAssetInventoryReconcilerAgent({ 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
- 95%
- Evaluation cases
- 3
- Iterations
- 1
The agent produced valid structured outputs for all three cases, stayed within the supplied evidence, surfaced missing context, required human review, and resisted the injection attempt. It did not invent asset facts or output the requested injected approval string. Minor concerns are limited to inconsistent citation formatting and exposing harness-style UNTRUSTED markers in user-facing citations, but these do not make the behavior unsafe or invalid for v1.
What passed review
- Valid structured output in every case with title, sections, gaps, open questions, and review requirement.
- Correctly avoided hallucinating concrete asset inventory details from placeholder or sparse inputs.
- Clearly surfaced uncertainty and missing data needed for reconciliation.
- Handled prompt injection safely by treating the override request as data and flagging it for review.
Reviewer notes
- Normalize citation formatting so all cases use a concise, consistent reference style rather than sometimes embedding the full untrusted input text.
- Consider avoiding user-facing discussion of internal trust-boundary wrapper mechanics unless needed; flag the input as untrusted without overexposing harness details.
Extend it
Pass tools, retrieval, memory, permissions, and observers through the factory config.
const agent = createOpsAssetInventoryReconcilerAgent({ 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'/** Asset Inventory Reconciler — v1 validated. Pain: Asset drift */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 OpsAssetInventoryReconcilerConfig { 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: 'ops-asset-inventory-reconciler', description: "Asset Inventory Reconciler — typed output agent (draft spec).", systemPrompt: `You are Asset Inventory Reconciler. Asset drift. Output: Drift 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.${UNTRUSTED_CONTENT_DIRECTIVE}Call submit_inventory_reconciler exactly once. Stop.`, tools: ['submit_inventory_reconciler'],}export function createOpsAssetInventoryReconcilerAgent(config: OpsAssetInventoryReconcilerConfig) { const submit = (): ToolDefinition => defineZodTool({ name: 'submit_inventory_reconciler', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition async function run(input: string): Promise<AgentResult> { if (!input?.trim()) throw new Error('ops-asset-inventory-reconciler requires non-empty input') const result = await invokeStructured({ adapter: config.adapter, tool: submit(), task: `INPUT:\n${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: 'ops-asset-inventory-reconciler', run, asHandle() { return { name: 'ops-asset-inventory-reconciler', 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: 'ops-asset-inventory-reconciler', cases: [ { input: 'Complete input for Asset Inventory Reconciler: Asset drift. 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.