vvzvlad/gitmost
1
0
Fork 0
Code Issues 22 Pull Requests 10 Packages Projects Releases Activity

feat(config): lower MCP timeouts and raise JSON body limit

Browse Source 
Reduce the default external-MCP silence timeout from 5 min to 1 min and the
overall call timeout from 15 min to 2 min. Update related tests and comments.
Increase Fastify's JSON body limit to 25 MiB (configurable via
HTTP_JSON_BODY_LIMIT) to accommodate large AI‑chat payloads.

BREAKING CHANGE: shorter MCP timeouts may abort long‑running tool calls that
previously succeeded with the older defaults.
This commit is contained in:
agent_vscode
2026-07-07 04:19:26 +03:00
parent f0afb2d729
commit 888c87f984
5 changed files with 45 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.env.example
+13 -5
View File
@@ -191,16 +191,24 @@ MCP_DOCMOST_PASSWORD=
# Silence timeout (ms) for EXTERNAL-MCP transport ONLY (not the chat provider).
# Tighter than AI_STREAM_TIMEOUT_MS so a byte-silent/hung MCP server is broken in
# ~5 min instead of 15. Note it also cuts a legitimately long but byte-silent
# ~1 min instead of 15. Note it also cuts a legitimately long but byte-silent
# single tool call (a slow crawl that emits nothing until done) and an SSE
# transport idling >5 min BETWEEN tool calls. Default 300000 (5 min).
# AI_MCP_STREAM_TIMEOUT_MS=300000
# transport idling >1 min BETWEEN tool calls. Default 60000 (1 min).
# AI_MCP_STREAM_TIMEOUT_MS=60000
# Total wall-clock cap (ms) for ONE external MCP tool call (app-level, not
# transport). Aborts a tool that keeps the socket warm (SSE heartbeats / trickle)
# but never returns a result — which the silence timeout above never breaks.
# Default 900000 (15 min).
# AI_MCP_CALL_TIMEOUT_MS=900000
# Default 120000 (2 min).
# AI_MCP_CALL_TIMEOUT_MS=120000
# Max JSON/urlencoded request body size (bytes). Fastify's 1 MiB default is too
# small for a long AI-chat research turn: the client resends the FULL message
# history (every tool call + search result) on each turn, so a deep conversation's
# POST to /api/ai-chat/stream can be several MB and would otherwise be rejected
# with FST_ERR_CTP_BODY_TOO_LARGE (413). Does NOT affect multipart file uploads
# (see FILE_UPLOAD_SIZE_LIMIT). Default 26214400 (25 MiB).
# HTTP_JSON_BODY_LIMIT=26214400
# Deferred tool loading for the in-app AI chat (#332). Default ON: the agent sees
# a compact <tool_catalog> and only CORE tools + a loadTools meta-tool are active
apps/server/src/core/ai-chat/external-mcp/mcp-call-timeout.spec.ts
+8 -8
View File
@@ -181,25 +181,25 @@ describe('mcp timeout env helpers', () => {
else process.env.AI_MCP_CALL_TIMEOUT_MS = ORIG_CALL;
});
it('mcpStreamTimeoutMs defaults to 5 min and honors a positive override', () => {
it('mcpStreamTimeoutMs defaults to 1 min and honors a positive override', () => {
delete process.env.AI_MCP_STREAM_TIMEOUT_MS;
expect(mcpStreamTimeoutMs()).toBe(300_000);
process.env.AI_MCP_STREAM_TIMEOUT_MS = '60000';
expect(mcpStreamTimeoutMs()).toBe(60_000);
process.env.AI_MCP_STREAM_TIMEOUT_MS = '90000';
expect(mcpStreamTimeoutMs()).toBe(90_000);
for (const bad of ['0', '-1', 'x', '']) {
process.env.AI_MCP_STREAM_TIMEOUT_MS = bad;
expect(mcpStreamTimeoutMs()).toBe(300_000);
expect(mcpStreamTimeoutMs()).toBe(60_000);
}
});
it('mcpCallTimeoutMs defaults to 15 min and honors a positive override', () => {
it('mcpCallTimeoutMs defaults to 2 min and honors a positive override', () => {
delete process.env.AI_MCP_CALL_TIMEOUT_MS;
expect(mcpCallTimeoutMs()).toBe(900_000);
process.env.AI_MCP_CALL_TIMEOUT_MS = '120000';
expect(mcpCallTimeoutMs()).toBe(120_000);
process.env.AI_MCP_CALL_TIMEOUT_MS = '180000';
expect(mcpCallTimeoutMs()).toBe(180_000);
for (const bad of ['0', '-1', 'x', '']) {
process.env.AI_MCP_CALL_TIMEOUT_MS = bad;
expect(mcpCallTimeoutMs()).toBe(900_000);
expect(mcpCallTimeoutMs()).toBe(120_000);
}
});
});
apps/server/src/core/ai-chat/external-mcp/mcp-clients.service.ts
+3 -3
View File
@@ -523,12 +523,12 @@ export function validateResolvedAddresses(addrs: readonly LookupAddress[]): {
*/
function buildPinnedDispatcher(): Agent {
// External-MCP traffic uses a DEDICATED, shorter silence timeout
// (`AI_MCP_STREAM_TIMEOUT_MS`, default 5 min) — deliberately tighter than the
// (`AI_MCP_STREAM_TIMEOUT_MS`, default 1 min) — deliberately tighter than the
// chat provider's 15-min `streamTimeoutMs()` — so a byte-silent/hung MCP
// upstream is broken in ~5 min instead of 15. We keep the keep-alive options
// upstream is broken in ~1 min instead of 15. We keep the keep-alive options
// from `streamingDispatcherOptions()` but OVERRIDE headers/body timeouts.
// Accepted trade-off: a legitimately long but byte-silent single tool call,
// and an SSE transport idling >5 min BETWEEN tool calls, are also cut here; the
// and an SSE transport idling >1 min BETWEEN tool calls, are also cut here; the
// per-call total cap (wrapToolsWithCallTimeout, `AI_MCP_CALL_TIMEOUT_MS`) is the
// complementary guard for chatty-but-stuck calls that keep the socket warm yet
// never return.
apps/server/src/integrations/ai/ai-streaming-fetch.ts
+7 -7
View File
@@ -123,19 +123,19 @@ export function streamKeepAliveMs(): number {
return positiveEnv('AI_STREAM_KEEPALIVE_MS', DEFAULT_STREAM_KEEPALIVE_MS);
}
/** Default SILENCE timeout for EXTERNAL-MCP transport (5 min). */
const DEFAULT_MCP_STREAM_TIMEOUT_MS = 300_000;
/** Default SILENCE timeout for EXTERNAL-MCP transport (1 min). */
const DEFAULT_MCP_STREAM_TIMEOUT_MS = 60_000;
/** Default total wall-clock cap for ONE external MCP tool call (15 min). */
const DEFAULT_MCP_CALL_TIMEOUT_MS = 900_000;
/** Default total wall-clock cap for ONE external MCP tool call (2 min). */
const DEFAULT_MCP_CALL_TIMEOUT_MS = 120_000;
/**
* SILENCE timeout (ms) for EXTERNAL-MCP transport ONLY. Override with
* `AI_MCP_STREAM_TIMEOUT_MS`; a missing/invalid/non-positive value falls back to
* {@link DEFAULT_MCP_STREAM_TIMEOUT_MS} (5 min).
* {@link DEFAULT_MCP_STREAM_TIMEOUT_MS} (1 min).
*
* Deliberately tighter than the chat provider's {@link streamTimeoutMs} (15 min)
* so a byte-silent/hung MCP upstream is broken in ~5 min instead of 15. This is
* so a byte-silent/hung MCP upstream is broken in ~1 min instead of 15. This is
* the undici `headersTimeout`/`bodyTimeout` for the external-MCP dispatcher only
* — it must NOT change the chat provider, which legitimately needs 15 min between
* reasoning chunks (#175).
@@ -153,7 +153,7 @@ export function mcpStreamTimeoutMs(): number {
/**
* Total wall-clock cap (ms) for ONE external MCP tool call — APP-LEVEL, not
* transport. Override with `AI_MCP_CALL_TIMEOUT_MS`; a missing/invalid/
* non-positive value falls back to {@link DEFAULT_MCP_CALL_TIMEOUT_MS} (15 min).
* non-positive value falls back to {@link DEFAULT_MCP_CALL_TIMEOUT_MS} (2 min).
*
* Catches a tool that keeps the connection warm (SSE heartbeats / trickle) but
* never returns a result — which the transport silence timeout
apps/server/src/main.ts
+14
View File
@@ -21,10 +21,24 @@ import { recordHttpResponse } from './integrations/metrics/http-metrics.hook';
import { startMetricsServer } from './integrations/metrics/metrics.server';
async function bootstrap() {
// Fastify JSON body cap. Fastify defaults to 1 MiB, which a long AI-chat
// research turn exceeds: the client resends the FULL message history (every
// tool call + search result) on each turn, so a deep conversation's POST to
// /api/ai-chat/stream can be several MB and would otherwise be rejected with
// FST_ERR_CTP_BODY_TOO_LARGE (413). Raise the cap; override with
// HTTP_JSON_BODY_LIMIT (bytes). A missing/invalid/non-positive value keeps the
// 25 MiB default. Multipart uploads are unaffected (their own @fastify/multipart
// limits apply); this only bounds JSON/urlencoded request bodies.
const bodyLimitEnv = Number(process.env.HTTP_JSON_BODY_LIMIT);
const bodyLimit =
Number.isFinite(bodyLimitEnv) && bodyLimitEnv > 0
? bodyLimitEnv
: 25 * 1024 * 1024;
const app = await NestFactory.create<NestFastifyApplication>(
AppModule,
new FastifyAdapter({
trustProxy: resolveTrustProxy(process.env.TRUST_PROXY),
bodyLimit,
routerOptions: {
maxParamLength: 1000,
ignoreTrailingSlash: true,