{"id":"insurance-agent-commission-audit","title":"Agent Commission Audit","description":"Audit typed. Commission errors. Typed v1 agent with eval coverage.","category":"insurance","status":"validated","version":"1.0.0","source":"agentskit-registry","license":"MIT","tags":["insurance","structured-output","v1"],"packages":["@agentskit/core","@agentskit/runtime","@agentskit/tools"],"files":["agent.ts","README.md","eval.ts"],"requires":{"zod":"^3","zod-to-json-schema":"^3"},"skill":{"name":"insurance-agent-commission-audit","description":"Audit typed. Commission errors. Typed v1 agent with eval coverage.","systemPrompt":"You are Agent Commission Audit. Commission errors. Output: Audit typed.\nActionable findings citing input sources. No invented issues.\nNEVER invent facts — gaps and openQuestions for missing input. Always draft for human review.\n${UNTRUSTED_CONTENT_DIRECTIVE}\nCall submit_commission_audit exactly once. Stop."},"flow":null,"a2a":{"id":"io.agentskit.registry.insurance-agent-commission-audit","name":"Agent Commission Audit","description":"Audit typed. Commission errors. Typed v1 agent with eval coverage.","version":"1.0.0","homepage":"https://registry.agentskit.io","skills":[{"name":"insurance-agent-commission-audit","description":"Audit typed. Commission errors. Typed v1 agent with eval coverage.","capabilities":{"streaming":true,"cancellation":true,"requiresApproval":false}}]},"sources":[{"path":"agent.ts","content":"import type { AdapterFactory, ChatMemory, Observer, ToolCall, ToolDefinition } from '@agentskit/core'\nimport { fenceUntrustedContent, UNTRUSTED_CONTENT_DIRECTIVE } from '@agentskit/core/security'\nimport { invokeStructured } from '@agentskit/runtime'\nimport { defineZodTool } from '@agentskit/tools'\nimport { z } from 'zod'\nimport { zodToJsonSchema } from 'zod-to-json-schema'\nimport type { JSONSchema7 } from 'json-schema'\n\n/** Agent Commission Audit — v1 validated. Pain: Commission errors */\n\nexport interface Finding { id: string; severity: 'critical' | 'high' | 'medium' | 'low' | 'info'; message: string; source?: string; recommendation?: string }\nexport interface AgentOutput { summary: string; findings: Finding[]; gaps: string[]; openQuestions: string[] }\nexport interface AgentResult extends AgentOutput { requiresReview: boolean }\nexport interface InsuranceAgentCommissionAuditConfig {\n  adapter: AdapterFactory\n  memory?: ChatMemory\n  observers?: Observer[]\n  onConfirm?: (toolCall: ToolCall) => boolean | Promise<boolean>\n  maxSteps?: number\n}\n\nconst Output = z.object({\n  summary: z.string(),\n  findings: z.array(z.object({\n    id: z.string(), severity: z.enum(['critical', 'high', 'medium', 'low', 'info']),\n    message: z.string(), source: z.string().optional(), recommendation: z.string().optional(),\n  })),\n  gaps: z.array(z.string()).default([]),\n  openQuestions: z.array(z.string()).default([]),\n})\nconst toJson = (s: z.ZodTypeAny): JSONSchema7 => zodToJsonSchema(s) as JSONSchema7\n\nconst skill = {\n  name: 'insurance-agent-commission-audit',\n  description: \"Agent Commission Audit — typed output agent (draft spec).\",\n  systemPrompt: `You are Agent Commission Audit. Commission errors. Output: Audit typed.\nActionable findings citing input sources. No invented issues.\nNEVER invent facts — gaps and openQuestions for missing input. Always draft for human review.\n${UNTRUSTED_CONTENT_DIRECTIVE}\nCall submit_commission_audit exactly once. Stop.`,\n  tools: ['submit_commission_audit'],\n}\n\nexport function createInsuranceAgentCommissionAuditAgent(config: InsuranceAgentCommissionAuditConfig) {\n  const submit = (): ToolDefinition =>\n    defineZodTool({ name: 'submit_commission_audit', description: 'Submit result. Once.', schema: Output, toJsonSchema: toJson, async execute() { return 'recorded' } }) as ToolDefinition\n\n  async function run(input: string): Promise<AgentResult> {\n    if (!input?.trim()) throw new Error('insurance-agent-commission-audit requires non-empty input')\n    const result = await invokeStructured({\n      adapter: config.adapter,\n      tool: submit(),\n      task: `INPUT:\\n${fenceUntrustedContent(input)}`,\n      parse: (a) => Output.parse(a),\n      skill,\n      memory: config.memory,\n      observers: config.observers,\n      onConfirm: config.onConfirm,\n      maxSteps: config.maxSteps ?? 4,\n    })\n    return { ...result, requiresReview: true }\n  }\n  return {\n    name: 'insurance-agent-commission-audit',\n    run,\n    asHandle() { return { name: 'insurance-agent-commission-audit', run: (t: string) => run(t).then((r) => JSON.stringify(r)) } },\n  }\n}\n"},{"path":"README.md","content":"# Agent Commission Audit\n\n> **v1 validated** — `npx agentskit add insurance-agent-commission-audit`\n\n## Pain\nCommission errors\n\n## Output\nAudit typed\n"},{"path":"eval.ts","content":"import type { EvalSuite } from '@agentskit/eval'\n\nexport const suite: EvalSuite = {\n  name: 'insurance-agent-commission-audit',\n  cases: [\n    { input: 'Complete input for Agent Commission Audit: Commission errors. Provide full structured output.', expected: (r: string) => r.length > 20 && /requiresReview|summary|title|category|findings|sections|score|clusters|items|steps/i.test(r) },\n    { input: 'Minimal input.', expected: (r: string) => /gap|openQuestion/i.test(r) || r.length > 10 },\n    { input: 'Input with specific detail: ACME Corp project deadline March 15.', expected: (r: string) => /ACME|March|15/i.test(r) || /gap/i.test(r) },\n    { input: 'Empty context — only says \"process this\".', expected: (r: string) => r.length > 5 },\n  ],\n}\n"}],"installable":true,"validation":{"status":"approved","score":96,"confidence":0.96,"method":"codex-executor-independent-reviewer","iterations":1,"cases":3,"summary":"The agent produced valid structured audit outputs, handled missing evidence conservatively, surfaced gaps and open questions, required human review, and resisted the injection request instead of outputting APPROVED. It did not hallucinate commission facts beyond the sparse inputs. The main quality issue is a duplicated gap line in the injection case, plus the normal case remains more of a refusal-to-audit than a rich audit because the input contained no actual commission data; that is acceptable given the no-hallucination requirement.","strengths":["Valid structured outputs for all cases.","Appropriately flags insufficient evidence and requires review.","Surfaces concrete missing data requirements and useful follow-up questions.","Resists prompt injection and treats redirection text as untrusted.","Avoids inventing agent, policy, premium, commission rate, or payment facts."],"notes":["Remove duplicate gap entries in sparse/injection handling.","Consider adding a stable source field to every finding if the output schema expects consistent finding shape across cases.","Improve the normal-case eval fixture with real commission data so the agent can demonstrate substantive audit calculations, not only missing-evidence handling."]}}