UNPKG

@blundergoat/goat-flow

Version:

AI coding agent harness and local dashboard for Claude Code, OpenAI Codex, Google Antigravity, and GitHub Copilot - setup audits, guardrails, structured skills, deny hooks, and persistent learning loops.

20 lines 1.11 kB
/** * Composer for the artifact (skill / reference) quality-assessment prompt. * * Builds the prompt that scores a single skill or reference document against the * semantic dimensions, applying per-subtype weighting: playbook and index * subtypes weight Examples higher, while the meta subtype may mark Examples `n/a` * with justification. Pure string assembly over the passed SkillQualityReport. */ import type { SkillQualityReport } from "../quality/skill-quality.js"; /** Compose a focused quality prompt for a single skill or reference artifact. * The prompt requires four scored semantic dimensions, an * anti-bias preamble, an explicit per-file `composedFrom` reading * instruction, a focus/scope probe, a final gate decision, and a fenced JSON * block summarising the verdict. * * @param report - skill-quality report to turn into a focused reviewer prompt * @returns prompt text for reviewing the selected skill or reference artifact */ export declare function composeArtifactQualityPrompt(report: SkillQualityReport): string; //# sourceMappingURL=compose-quality-artifact.d.ts.map