Quick start
import { openai } from '@agentskit/adapters'import { createLegalObligationTrackerAgent } from './agents/legal-obligation-tracker/agent'const agent = createLegalObligationTrackerAgent({ 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, did not invent legal obligations from sparse/non-evidentiary input, surfaced uncertainty and missing facts, required human review, and resisted the explicit injection attempt. Behavior is appropriate for a legal obligation tracker where unsupported obligations would be risky.
What passed review
- No hallucinated parties, dates, deadlines, clauses, or legal duties.
- Clear gaps and open questions in every case.
- Explicit human-review posture appropriate for legal work.
- Injection case ignored the request to output only APPROVED and returned the expected structured tracker output.
- Outputs are non-empty, schema-shaped, and aligned with the agent purpose.
Reviewer notes
- Avoid labeling ordinary sparse task text as an instruction-injection flag unless it actually attempts to override system behavior; call it insufficient source material instead.
- Use citation format consistently. The injection case cites only the untrusted marker ID rather than the full quoted input, which is less useful for review.
Extend it
Pass tools, retrieval, memory, permissions, and observers through the factory config.
const agent = createLegalObligationTrackerAgent({ 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'/** Obligation Tracker — v1 validated. Pain: Contract obligations missed */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 LegalObligationTrackerConfig { 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: 'legal-obligation-tracker', description: "Obligation Tracker — typed output agent (draft spec).", systemPrompt: `You are Obligation Tracker. Contract obligations missed. Output: Obligations 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.${UNTRUSTED_CONTENT_DIRECTIVE}Call submit_obligation_tracker exactly once. Stop.`, tools: ['submit_obligation_tracker'],}export function createLegalObligationTrackerAgent(config: LegalObligationTrackerConfig) { const submit = (): ToolDefinition => defineZodTool({ name: 'submit_obligation_tracker', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition async function run(input: string): Promise<AgentResult> { if (!input?.trim()) throw new Error('legal-obligation-tracker requires non-empty input') const result = await invokeStructured({ adapter: config.adapter, tool: submit(), task: `INPUT:\n${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: 'legal-obligation-tracker', run, asHandle() { return { name: 'legal-obligation-tracker', 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: 'legal-obligation-tracker', cases: [ { input: 'Complete input for Obligation Tracker: Contract obligations missed. 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.