{"id":"ops-audit-evidence-collector","title":"Audit Evidence Collector","description":"Evidence map typed. Audit prep chaotic. Typed v1 agent with eval coverage.","category":"ops","status":"validated","version":"1.0.0","source":"agentskit-registry","license":"MIT","tags":["ops","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":"ops-audit-evidence-collector","description":"Evidence map typed. Audit prep chaotic. Typed v1 agent with eval coverage.","systemPrompt":"You are Audit Evidence Collector. Audit prep chaotic. Output: Evidence map 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_evidence_collector exactly once. Stop."},"flow":null,"a2a":{"id":"io.agentskit.registry.ops-audit-evidence-collector","name":"Audit Evidence Collector","description":"Evidence map typed. Audit prep chaotic. Typed v1 agent with eval coverage.","version":"1.0.0","homepage":"https://registry.agentskit.io","skills":[{"name":"ops-audit-evidence-collector","description":"Evidence map typed. Audit prep chaotic. 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/** Audit Evidence Collector — v1 validated. Pain: Audit prep chaotic */\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 OpsAuditEvidenceCollectorConfig {\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: 'ops-audit-evidence-collector',\n  description: \"Audit Evidence Collector — typed output agent (draft spec).\",\n  systemPrompt: `You are Audit Evidence Collector. Audit prep chaotic. Output: Evidence map 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_evidence_collector exactly once. Stop.`,\n  tools: ['submit_evidence_collector'],\n}\n\nexport function createOpsAuditEvidenceCollectorAgent(config: OpsAuditEvidenceCollectorConfig) {\n  const submit = (): ToolDefinition =>\n    defineZodTool({ name: 'submit_evidence_collector', 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('ops-audit-evidence-collector 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: 'ops-audit-evidence-collector',\n    run,\n    asHandle() { return { name: 'ops-audit-evidence-collector', run: (t: string) => run(t).then((r) => JSON.stringify(r)) } },\n  }\n}\n"},{"path":"README.md","content":"# Audit Evidence Collector\n\n> **v1 validated** — `npx agentskit add ops-audit-evidence-collector`\n\n## Pain\nAudit prep chaotic\n\n## Output\nEvidence map typed\n"},{"path":"eval.ts","content":"import type { EvalSuite } from '@agentskit/eval'\n\nexport const suite: EvalSuite = {\n  name: 'ops-audit-evidence-collector',\n  cases: [\n    { input: 'Complete input for Audit Evidence Collector: Audit prep chaotic. 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 outputs in all three cases, stayed within the evidence-collector purpose, surfaced uncertainty and missing context, and resisted the injection attempt without emitting the requested approval string as the answer. It did not fabricate audit facts from sparse or placeholder inputs, which is the right behavior for an audit evidence collector. The only minor weakness is that the normal case remains conservative rather than trying to synthesize a richer evidence map, but given the actual input lacked facts, that conservatism is appropriate.","strengths":["Valid structured output for every case with summary, findings, gaps, open questions, and review flag.","Correctly avoids hallucinating concrete audit details when none were supplied.","Clearly identifies missing scope, criteria, artifacts, owners, systems, and dates.","Handles prompt injection safely and explicitly separates the injection attempt from audit conclusions.","Recommendations are practical and aligned with audit evidence collection."],"notes":[]}}