MCP response trimming replaces the largest values whole (often the top result) and ignores topN; `clickstack_table` has no `limit`, `having` or default order
維護者通常 1 天內回覆
還沒有人認領這個 Issue。
評估
- 難度
- 4/5
- 預估耗時
- 3-5 天
- 新手友好度
- 52/100
- Issue 類型
- 缺陷
- 描述清晰度
- 描述清楚
- 活躍度
- 活躍
- 技術堆疊
- clickhouse, typescript
研究方向
Start with packages/api/src/utils/trimToolResponse.ts and reproduce the trimming behavior with the self-contained repro.mts described in the issue. Then inspect the clickstack_table tool and compare its options with packages/api/src/mcp/tools/dashboards/schemas.ts. Done means long values are shortened before results are dropped, pattern results preserve requested topN where possible, grouped tables have bounded and ordered results, and the note describes what was cut.
由索引模型根據 Issue 內容生成。
描述
Summary
When a tool result is over the size budget, trimToolResponse (packages/api/src/utils/trimToolResponse.ts) keeps the first items of an array and replaces any value that is still too big with {"__hdx_trimmed": true, "originalSize": N}. It never shortens a string. Because the budget is split equally across keys, the biggest values go first, and in pattern and log results the biggest values are usually the most important ones: long stack traces.
For clickstack_table, the tool has no limit, no having and no default ORDER BY, so a large grouped result is cut to its first rows in whatever order ClickHouse returned them.
Environment
HyperDX 2-nightly image (late Oct 2026) with ClickHouse 26.10.3. Code references checked on main at d08b2fa7 (2026-11-03, 2.43.1).
Steps to reproduce
clickstack_event_patternswithtopN: 30on a service whose top errors are multi-line stack traces.- The response has 10 patterns, and the top ones look like
{"rank":1,"pattern":{"__hdx_trimmed":true,"originalSize":7443},"whereSnippet":{"__hdx_trimmed":true,…},"samples":[{"__hdx_trimmed":true,…}]}.
Self-contained (repro.mts below; runs the real trimToolResponse from main on an event_patterns-shaped object with 3 long and 27 short patterns):
isTrimmed=true; topN requested=30, patterns returned=10
rank 1: pattern={"__hdx_trimmed":true,"originalSize":4736} whereSnippet={"__hdx_trimmed":true,…} samples[0]={"__hdx_trimmed":true,…}
rank 2: … same
rank 3: … same
rank 4: pattern="Short INFO line 0 took <*> ms" (complete)
clickstack_tablewith agroupBythat yields a few thousand groups and noorderBy: the response holds the first rows of an unordered result (one run: 2,638 rows in, about 75 out), plus "Result was trimmed".
Expected vs actual
| Expected | Actual | |
|---|---|---|
| Over-budget pattern result | the top N patterns, with long strings shortened (…(+6,900 chars)) and fewer samples |
10 patterns; the top 3 replaced by markers |
| Large grouped table | the top rows by the first select (or by orderBy), up to a limit |
the first rows in arbitrary order |
Impact
In 4 agent investigations on production-size data: the most frequent error pattern of one service (99.5% of its ERROR lines) came back as a marker in 3 of 4 calls; topN 15–30 returned 10 every time; 66 KB and 81 KB responses still overflowed the client; and table calls returned 89–12,194 unordered groups that were then cut.
Proposed fix
- In
trimToolResponse, shorten long strings (e.g. to 500–2,000 chars with a suffix saying how much was cut) before replacing values or dropping array items. - For pattern results, drop
samples(keep 1) before dropping patterns, and keep the requestedtopNwhen it fits after that. - Add
limit(default about 50) andhavingtoclickstack_table, as dashboard tiles already have (packages/api/src/mcp/tools/dashboards/schemas.ts:839,:947), and defaultORDER BYto the first select descending whengroupByis set. - Say in the
notewhat was cut (e.g. "20 patterns dropped, 3 strings shortened").
Related
- #2337: shared MCP helpers and error hints
repro.mts
// Repro for trimToolResponse (packages/api/src/utils/trimToolResponse.ts) on an
// event_patterns-shaped result: long values are replaced, not shortened, and the
// array is cut to 10 items regardless of topN.
// Run from <hyperdx>/packages/api: npx tsx <this file> ./src/utils/trimToolResponse.ts
import { pathToFileURL } from 'node:url';
const { trimToolResponse } = await import(pathToFileURL(process.argv[2]).href);
const stack = (n: number) => `java.lang.NullPointerException: Cannot invoke "Catalog.items()" (case ${n})\n` +
Array.from({ length: 90 }, (_, i) => `\tat com.example.svc.Layer${i}.call(Layer${i}.java:${i + 10})`).join('\n'); // ~7.4 KB
const pattern = (rank: number, text: string, count: number) => ({
rank, id: `p${rank}`, pattern: text, estimatedCount: count, sampleCount: count, shareOfTotal: 0,
whereSnippet: `Body:"${text.slice(0, 2000)}"`, samples: Array.from({ length: 5 }, () => text),
});
const patterns = [
...[0, 1, 2].map(i => pattern(i + 1, stack(i), 9950 - i)), // the 3 most frequent: long stack traces
...Array.from({ length: 27 }, (_, i) => pattern(i + 4, `Short INFO line ${i} took <*> ms`, 30 - i)),
];
const output = { summary: { totalCount: 100000, patternsReturned: patterns.length }, patterns, usage: 'shareOfTotal is …' };
const { data, isTrimmed } = trimToolResponse(output);
console.log(`isTrimmed=${isTrimmed}; topN requested=30, patterns returned=${data.patterns.length}`);
for (const p of data.patterns.slice(0, 4)) {
const show = (v: unknown) => (typeof v === 'string' ? `"${v.slice(0, 40)}…" (${v.length} chars)` : JSON.stringify(v));
console.log(`rank ${p.rank}: pattern=${show(p.pattern)} whereSnippet=${show(p.whereSnippet)} samples[0]=${show(Array.isArray(p.samples) ? p.samples[0] : p.samples)}`);
}
- 主要語言
- TypeScript
- 星號
- 10.1k
- 分支
- 527
- 平均合併
- 1 天 20 小時
- 30 天內合併 PR
- 186
環境準備
- 提供 Dockerfile 或 Docker Compose 檔案
- 有 Pull Request 範本
- 閱讀貢獻指南
從這裡開始
- 先讀完整個 Issue,再讀專案的貢獻指南。
- 在 Issue 下留言說明你要接手 —— 這能避免兩個人做同樣的事。
- Fork 儲存庫,在一個分支上完成修改。
- 送出 Pull Request,並在描述裡引用這個 Issue 編號。
ClickHouse/ClickStack 的其他 Issue
-
external
難度 2/5 1-3 小時 新手友好度 72/100
ClickHouse/ClickStack#3761 ·
維護者通常 1 天內回覆
-
Metric chart with aggFn increase and a multi-column group-by fails: toString() gets several arguments可能已有人在做 @CG-5228 於 26 天前認領。 未關閉external
難度 2/5 1-3 小時 新手友好度 76/100
ClickHouse/ClickStack#3477 · 1 則留言 ·
維護者通常 1 天內回覆
-
Slack notifications for inline alerts fail when the chart-explorer link passes Slack's 3000-character block limit可能已有人在做 @Tyagiquamar 於 27 天前認領。 未關閉alerting bug external
難度 2/5 1-3 小時 新手友好度 76/100
ClickHouse/ClickStack#3452 ·
維護者通常 1 天內回覆
-
check-alerts worker: 13 error spans (mongodb.create, NamespaceExists) every minute because mongoose autoCreate is still on可能重新可做 關聯的 PR 已關閉且未合併。 未關閉bug external Needs Feedback
難度 2/5 1-3 小時 新手友好度 72/100
ClickHouse/ClickStack#3308 · 1 則留言 ·
維護者通常 1 天內回覆
-
count_distinct aggFn on metric tiles ignores valueExpression, always counts distinct metric Values可能已有人在做 @Harshul1484 於 54 天前認領。 未關閉
難度 2/5 1-3 小時 新手友好度 72/100
ClickHouse/ClickStack#2311 ·
維護者通常 1 天內回覆
查看 ClickHouse/ClickStack 的全部 Issue
相似的 Issue
-
Add: GTrekker未關閉check:passed streams:add
難度 1/5 1 小時以內 新手友好度 72/100
維護者通常 1 天內回覆
-
area: ui Bug Payload OSS
難度 1/5 1 小時以內 新手友好度 88/100
payloadcms/payload#19839 ·
維護者通常 1 天內回覆
-
難度 2/5 1-3 小時 新手友好度 85/100
維護者通常 1 天內回覆
-
fix(auth): add the email OTP verification step to Pollar login可能已有人在做 @akansnoro 今天認領。 未關閉Stellar Wave
難度 2/5 1-3 小時 新手友好度 74/100
safetrustcr/dApp-SafeTrust#584 · 10 則留言 ·
維護者通常 1 天內回覆
-
The manager preset saves launch definitions without the required file and digest pair可能已有人在做 關聯的 PR 仍在進行中或已合併。 未關閉bug
難度 2/5 1-3 小時 新手友好度 80/100
維護者通常 1 天內回覆