Skip to main content

Interface: EmergentJudgeConfig

Defined in: packages/agentos/src/cognition/emergent/EmergentJudge.ts:101

Configuration for the EmergentJudge.

All LLM interaction is abstracted behind the generateText callback, making the judge model-agnostic and easily testable with mocks.

Properties​

generateText()​

generateText: (model, prompt) => Promise<string>

Defined in: packages/agentos/src/cognition/emergent/EmergentJudge.ts:126

Callback that invokes an LLM to generate text from a prompt. The judge calls this for creation reviews and promotion panels.

Parameters​

model​

string

The model ID to use for generation.

prompt​

string

The full prompt string to send to the LLM.

Returns​

Promise<string>

The raw text response from the LLM.

generateTextWithSystem()?​

optional generateTextWithSystem: (model, system, user) => Promise<string>

Defined in: packages/agentos/src/cognition/emergent/EmergentJudge.ts:146

Optional structured callback that receives a stable system prefix and a candidate-specific user payload separately. When supplied, the judge prefers this path over generateText so hosts can attach provider-level prompt caching (e.g. Anthropic cache_control: ephemeral or OpenAI automatic prefix cache) to the shared rubric. A 10-20 call run on Anthropic sees ~25% judge cost reduction once the ~500-token rubric hits the cache on call 2+.

Hosts that do not care about caching may omit this field; the judge falls back to concatenating system + '\n\n' + user and calling the legacy generateText path, which preserves behavior exactly.

Parameters​

model​

string

Model ID to use for generation.

system​

string

Stable rubric text. Safe to mark cacheable.

user​

string

Candidate-specific payload that varies per call.

Returns​

Promise<string>

The raw text response from the LLM.

judgeModel​

judgeModel: string

Defined in: packages/agentos/src/cognition/emergent/EmergentJudge.ts:108

Model ID used for the single-pass creation review. Should be a fast, cost-efficient model since correctness is primarily validated through test cases.

Example​

"gpt-4o-mini"

promotionModel​

promotionModel: string

Defined in: packages/agentos/src/cognition/emergent/EmergentJudge.ts:116

Model ID used by both reviewers in the promotion panel. Should be a more capable model than judgeModel since promotion decisions are higher-stakes.

Example​

"gpt-4o"