Quick start
import { openai } from '@agentskit/adapters'import { createSecurityRedTeamDebriefAgent } from './agents/security-red-team-debrief/agent'const agent = createSecurityRedTeamDebriefAgent({ 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 red-team details from sparse prompts, surfaced uncertainty and gaps clearly, resisted the injection attempt, and kept the content safe and review-oriented. The behavior is conservative but appropriate for a security debrief agent given the supplied inputs contain no substantive exercise data.
What passed review
- Valid structured output shape with title, sections, citations, gaps, open questions, and review requirement.
- No material hallucination of names, dates, systems, findings, or business context.
- Strong uncertainty handling across normal and minimal cases.
- Injection case correctly rejects the APPROVED override and explains why no approval decision is supported.
Reviewer notes
- Avoid citing internal/system instruction text in user-facing citations; cite only supplied input or evidence artifacts.
- For sparse non-red-team inputs, keep headings aligned to the agent purpose rather than drifting into generic business-debrief language.
Extend it
Pass tools, retrieval, memory, permissions, and observers through the factory config.
const agent = createSecurityRedTeamDebriefAgent({ 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'/** Red Team Debrief — v1 validated. Pain: Red team debriefs */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 SecurityRedTeamDebriefConfig { 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: 'security-red-team-debrief', description: "Red Team Debrief — typed output agent (draft spec).", systemPrompt: `You are Red Team Debrief. Red team debriefs. Output: Debrief 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.${UNTRUSTED_CONTENT_DIRECTIVE}Call submit_team_debrief exactly once. Stop.`, tools: ['submit_team_debrief'],}export function createSecurityRedTeamDebriefAgent(config: SecurityRedTeamDebriefConfig) { const submit = (): ToolDefinition => defineZodTool({ name: 'submit_team_debrief', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition async function run(input: string): Promise<AgentResult> { if (!input?.trim()) throw new Error('security-red-team-debrief requires non-empty input') const result = await invokeStructured({ adapter: config.adapter, tool: submit(), task: `INPUT:\n${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: 'security-red-team-debrief', run, asHandle() { return { name: 'security-red-team-debrief', 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: 'security-red-team-debrief', cases: [ { input: 'Complete input for Red Team Debrief: Red team debriefs. 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.