fix: don't close a shared MCP transport under in-flight sibling calls

[bhaveshpatel640]

Summary

Session-expired MCP retry no longer closes the shared transport while a sibling callTool/listTools is still using it. That sibling used to fail with a non-retried close error.

AGE-2253 / INFOSEC-15 TF-06

Changes

• Detach the stale socket (pendingClose) and close it when inflight hits 0
• Skip reset if the expired handle is no longer the live connection

How was this tested?

pnpm --filter @truefoundry/trueforge-core test -- tests/core/mcp/remoteMcpServer.test.ts

Checklist

• I have read the contributing guidelines
• pnpm build, pnpm test, pnpm typecheck, pnpm lint:ci, and pnpm format:check pass locally
• Tests added/updated where it makes sense
• No hand-edits to generated code (packages/trueforge-sdk, python/trueforge_sdk, .github/fern/openapi/openapi.json, docs/openapi.json) — fork PRs omit SDK regen; maintainers regenerate after merge
• Docs / .env.example updated if configuration or behavior changed

---

Note

Medium Risk
Changes concurrent connection teardown and session-expiry retry in core remote MCP usage; behavior is narrow and covered by new tests, but mistakes could affect parallel tool calls or leak connections.

Overview
Fixes a race where session-expired retry on one concurrent callTool/listTools would close the shared remote MCP transport while another call still held it. That sibling then failed with "Connection closed", which is not treated as session-expired, so it was never retried.

RemoteMCP now tracks in-flight operations on a captured connection handle: on expiry it detaches the stale socket into pendingClose instead of closing immediately when siblings are active, reconnects and retries once, and closes the deferred socket only when inflight reaches zero. Stale expiry on an old handle no longer resets a newer live transport, and transport onClose only clears connected state when the closed socket is still the active one.

Tests simulate a held sibling call plus session expiry on a parallel call, including late expiry on the stale connection after retry.

^{Reviewed by Cursor Bugbot for commit 1dfd55c. Bugbot is set up for automated code reviews on this repo. Configure here.}

[changeset-bot]

🦋 Changeset detected

Latest commit: b57abe5

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 2 packages

Name	Type
@truefoundry/trueforge-core	Patch
@truefoundry/trueforge	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit faf9c05. Configure here.}