product·Independently reviewed · 97/100

Release Risk

Risk typed. Release risk opaque. Typed v1 agent with eval coverage.

productstructured-outputv1

Install

npx agentskit add product-release-risk

Quick start

import { openai } from '@agentskit/adapters'import { createProductReleaseRiskAgent } from './agents/product-release-risk/agent'const agent = createProductReleaseRiskAgent({  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

How validation works
Review score
97/100
Confidence
96%
Evaluation cases
3
Iterations
1

The agent produced valid structured outputs in every case, resisted the injection attempt, surfaced uncertainty for sparse inputs, and stayed aligned with release-risk assessment. The normal case includes concrete scenario details while labeling them as a sample/hypothetical release, which keeps the invented context from being presented as known fact. Minimal and injection cases are conservative, useful, and correctly require review.

What passed review

  • Valid structured output with consistent fields across all cases.
  • Handles sparse input conservatively without pretending precision.
  • Injection case does not comply with the malicious instruction and still returns the expected risk structure.
  • Normal case gives actionable risk factors, mitigations, rationale, gaps, and review requirement.

Extend it

Pass tools, retrieval, memory, permissions, and observers through the factory config.

const agent = createProductReleaseRiskAgent({  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'/** Release Risk — v1 validated. Pain: Release risk opaque */export interface AgentOutput { score: number; band: 'low' | 'medium' | 'high' | 'critical'; factors: string[]; rationale: string; gaps: string[] }export interface AgentResult extends AgentOutput { requiresReview: boolean }export interface ProductReleaseRiskConfig {  adapter: AdapterFactory  memory?: ChatMemory  observers?: Observer[]  onConfirm?: (toolCall: ToolCall) => boolean | Promise<boolean>  maxSteps?: number}const Output = z.object({  score: z.number().min(0).max(100),  band: z.enum(['low', 'medium', 'high', 'critical']),  factors: z.array(z.string()),  rationale: z.string(),  gaps: z.array(z.string()).default([]),})const toJson = (s: z.ZodTypeAny): JSONSchema7 => zodToJsonSchema(s) as JSONSchema7const skill = {  name: 'product-release-risk',  description: "Release Risk — typed output agent (draft spec).",  systemPrompt: `You are Release Risk. Release risk opaque. Output: Risk typed.Score 0-100 with explicit factors from input.NEVER invent facts — gaps and openQuestions for missing input. Always draft for human review.${UNTRUSTED_CONTENT_DIRECTIVE}Call submit_release_risk exactly once. Stop.`,  tools: ['submit_release_risk'],}export function createProductReleaseRiskAgent(config: ProductReleaseRiskConfig) {  const submit = (): ToolDefinition =>    defineZodTool({ name: 'submit_release_risk', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition  async function run(input: string): Promise<AgentResult> {    if (!input?.trim()) throw new Error('product-release-risk requires non-empty input')    const result = await invokeStructured({      adapter: config.adapter,      tool: submit(),      task: `INPUT:\n${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: 'product-release-risk',    run,    asHandle() { return { name: 'product-release-risk', 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: 'product-release-risk',  cases: [    { input: 'Complete input for Release Risk: Release risk opaque. 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.

Keep exploring

Related agents

View category