@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.
23 lines • 1.27 kB
TypeScript
/**
* Builds the compact, ranked list of learning-loop entries (footguns, lessons,
* patterns, decisions) used for bounded prompt retrieval - each bucket file is
* split into sections, summarized into a small fact with a byte-capped excerpt,
* and tagged with reference-validation results.
*
* The final ordering is deterministic by design (kind, then newest-first by
* updated/created date, then path, then discovery order) so prompt context is
* stable across runs and machines. The per-kind order offsets (0, 10k, 20k, 30k)
* are tiebreak seeds that keep entries from different buckets from interleaving
* unpredictably; they are not limits on entry count.
*/
import type { LearningLoopEntryFact, ReadonlyFS } from "../../types.js";
import type { LoadedConfig } from "../../config/types.js";
/**
* Extract compact learning-loop entries for bounded prompt retrieval.
*
* @param fs - filesystem adapter for the target project
* @param configState - loaded config with footgun, lesson, and decision paths
* @returns ordered compact entries suitable for prompt context selection
*/
export declare function extractLearningLoopEntries(fs: ReadonlyFS, configState: LoadedConfig): LearningLoopEntryFact[];
//# sourceMappingURL=learning-loop-entries.d.ts.map