بنقرة واحدة
iterative-retrieval
Pattern for progressively refining context retrieval to solve the subagent context problem
القائمة
Pattern for progressively refining context retrieval to solve the subagent context problem
Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills/commands/agents. v2.1 adds project-scoped instincts to prevent cross-project contamination.
Orchestrate building a brand-new feature end to end — research, plan, TDD implementation, review, and gated commit — by delegating each phase to the matching ECC agent. Use when adding a capability that does not exist yet.
Orchestrate bootstrapping a working MVP from a design or spec document — ingest the doc, plan thin vertical slices, scaffold the first end-to-end slice, then TDD-implement, review, and gated commit. Use to turn an SDD/PRD into a running starting point.
Orchestrate altering an existing, working feature to new desired behavior — update its tests to the new spec, change the implementation to match, review, and gated commit. Use when behavior is not broken but should be different.
Orchestrate fixing a bug — reproduce it as a failing regression test, fix to green, review, and gated commit — by delegating each phase to the matching ECC agent. Use when existing behavior is broken or wrong.
Shared orchestration engine for the orch-* skill family. Defines the gated Research-Plan-TDD-Review-Commit pipeline, the size classifier, the agent map, and the two human gates that the orch-* operation skills delegate to. Not usually invoked directly.
| name | iterative-retrieval |
| description | Pattern for progressively refining context retrieval to solve the subagent context problem |
解決多 agent 工作流程中的「上下文問題」,其中子 agents 在開始工作之前不知道需要什麼上下文。
子 agents 以有限上下文產生。它們不知道:
標準方法失敗:
一個漸進精煉上下文的 4 階段循環:
┌─────────────────────────────────────────────┐
│ │
│ ┌──────────┐ ┌──────────┐ │
│ │ DISPATCH │─────│ EVALUATE │ │
│ └──────────┘ └──────────┘ │
│ ▲ │ │
│ │ ▼ │
│ ┌──────────┐ ┌──────────┐ │
│ │ LOOP │─────│ REFINE │ │
│ └──────────┘ └──────────┘ │
│ │
│ 最多 3 個循環,然後繼續 │
└─────────────────────────────────────────────┘
初始廣泛查詢以收集候選檔案:
// 從高層意圖開始
const initialQuery = {
patterns: ['src/**/*.ts', 'lib/**/*.ts'],
keywords: ['authentication', 'user', 'session'],
excludes: ['*.test.ts', '*.spec.ts']
};
// 派遣到檢索 agent
const candidates = await retrieveFiles(initialQuery);
評估檢索內容的相關性:
function evaluateRelevance(files, task) {
return files.map(file => ({
path: file.path,
relevance: scoreRelevance(file.content, task),
reason: explainRelevance(file.content, task),
missingContext: identifyGaps(file.content, task)
}));
}
評分標準:
基於評估更新搜尋標準:
function refineQuery(evaluation, previousQuery) {
return {
// 新增在高相關性檔案中發現的新模式
patterns: [...previousQuery.patterns, ...extractPatterns(evaluation)],
// 新增在程式碼庫中找到的術語
keywords: [...previousQuery.keywords, ...extractKeywords(evaluation)],
// 排除確認不相關的路徑
excludes: [...previousQuery.excludes, ...evaluation
.filter(e => e.relevance < 0.2)
.map(e => e.path)
],
// 針對特定缺口
focusAreas: evaluation
.flatMap(e => e.missingContext)
.filter(unique)
};
}
以精煉標準重複(最多 3 個循環):
async function iterativeRetrieve(task, maxCycles = 3) {
let query = createInitialQuery(task);
let bestContext = [];
for (let cycle = 0; cycle < maxCycles; cycle++) {
const candidates = await retrieveFiles(query);
const evaluation = evaluateRelevance(candidates, task);
// 檢查是否有足夠上下文
const highRelevance = evaluation.filter(e => e.relevance >= 0.7);
if (highRelevance.length >= 3 && !hasCriticalGaps(evaluation)) {
return highRelevance;
}
// 精煉並繼續
query = refineQuery(evaluation, query);
bestContext = mergeContext(bestContext, highRelevance);
}
return bestContext;
}
任務:「修復認證 token 過期 bug」
循環 1:
DISPATCH:在 src/** 搜尋 "token"、"auth"、"expiry"
EVALUATE:找到 auth.ts (0.9)、tokens.ts (0.8)、user.ts (0.3)
REFINE:新增 "refresh"、"jwt" 關鍵字;排除 user.ts
循環 2:
DISPATCH:搜尋精煉術語
EVALUATE:找到 session-manager.ts (0.95)、jwt-utils.ts (0.85)
REFINE:足夠上下文(2 個高相關性檔案)
結果:auth.ts、tokens.ts、session-manager.ts、jwt-utils.ts
任務:「為 API 端點增加速率限制」
循環 1:
DISPATCH:在 routes/** 搜尋 "rate"、"limit"、"api"
EVALUATE:無匹配 - 程式碼庫使用 "throttle" 術語
REFINE:新增 "throttle"、"middleware" 關鍵字
循環 2:
DISPATCH:搜尋精煉術語
EVALUATE:找到 throttle.ts (0.9)、middleware/index.ts (0.7)
REFINE:需要路由器模式
循環 3:
DISPATCH:搜尋 "router"、"express" 模式
EVALUATE:找到 router-setup.ts (0.8)
REFINE:足夠上下文
結果:throttle.ts、middleware/index.ts、router-setup.ts
在 agent 提示中使用:
為此任務檢索上下文時:
1. 從廣泛關鍵字搜尋開始
2. 評估每個檔案的相關性(0-1 尺度)
3. 識別仍缺少的上下文
4. 精煉搜尋標準並重複(最多 3 個循環)
5. 回傳相關性 >= 0.7 的檔案
continuous-learning 技能 - 用於隨時間改進的模式~/.claude/agents/ 中的 Agent 定義