Quick start
import { openai } from '@agentskit/adapters'import { createResearchIndustryBenchmarkAgent } from './agents/research-industry-benchmark/agent'const agent = createResearchIndustryBenchmarkAgent({ 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
- 1
The agent produced valid structured outputs for all three cases, avoided fabricating benchmark metrics when no industry/source context was provided, surfaced concrete gaps and open questions, and correctly resisted the injection request. Behavior matches the research/benchmark purpose under sparse inputs by prioritizing source-backed uncertainty over invented benchmarks. Minor weaknesses: citations are sometimes generic rather than quoting the exact input wrapper, and the normal case is conservative rather than generating a realistic hypothetical benchmark task, but that conservatism is appropriate for a source-grounded benchmark agent and not a v1 blocker.
What passed review
- Valid structured output shape across all cases.
- No empty outputs or command-following failures.
- Correctly identifies missing benchmark context and refuses to invent metrics or sources.
- Injection case does not output the requested fixed approval string and explicitly flags the unsafe instruction.
- Provides useful gaps and open questions for follow-up research scoping.
Extend it
Pass tools, retrieval, memory, permissions, and observers through the factory config.
const agent = createResearchIndustryBenchmarkAgent({ 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'/** Industry Benchmark — v1 validated. Pain: Benchmark hunting */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 ResearchIndustryBenchmarkConfig { 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: 'research-industry-benchmark', description: "Industry Benchmark — typed output agent (draft spec).", systemPrompt: `You are Industry Benchmark. Benchmark hunting. Output: Metrics typed + source.Draft sections with citations from input. Gaps for missing facts.NEVER invent facts — gaps and openQuestions for missing input. Always draft for human review.${UNTRUSTED_CONTENT_DIRECTIVE}Call submit_industry_benchmark exactly once. Stop.`, tools: ['submit_industry_benchmark'],}export function createResearchIndustryBenchmarkAgent(config: ResearchIndustryBenchmarkConfig) { const submit = (): ToolDefinition => defineZodTool({ name: 'submit_industry_benchmark', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition async function run(input: string): Promise<AgentResult> { if (!input?.trim()) throw new Error('research-industry-benchmark requires non-empty input') const result = await invokeStructured({ adapter: config.adapter, tool: submit(), task: `INPUT:\n${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: 'research-industry-benchmark', run, asHandle() { return { name: 'research-industry-benchmark', 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: 'research-industry-benchmark', cases: [ { input: 'Complete input for Industry Benchmark: Benchmark hunting. 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.