Skip to content

Merge upstream main into fix/nsjail-no-cgroup-clone - #4

Merged
TomasPalsson merged 43 commits into
fix/nsjail-no-cgroup-clonefrom
chore/sync-upstream-main
Sep 25, 2026
Merged

TomasPalsson merged 43 commits into
fix/nsjail-no-cgroup-clonefrom
chore/sync-upstream-main

Conversation

@busla

@busla busla commented Sep 24, 2026

Copy link
Copy Markdown

Merges LibreChat-AI/code-interpreter main at 67d75d859aee891923c40cde1db073489d74a424 into fix/nsjail-no-cgroup-clone.

The merge conflict in service/src/file-server.ts, service/src/queue.ts, and service/src/workers.ts was resolved by retaining the fork’s MinIO credential and BullMQ namespace behavior while incorporating upstream’s file-object, cancellation, and programmatic execution changes.

Supersedes #3 and preserves the fix from #1.

danny-avila and others added 30 commits September 11, 2026 11:10
…t-AI#179)

* fix: distinguish retryable ledger contention from scope denials

* fix: preserve error classification through marker discovery and relay

* test: exercise classified denials through gateway configuration

* fix: honor bounded gateway retry hints during object downloads
…ic (LibreChat-AI#180)

* perf: reuse authorized input versions and make egress accounting atomic

* perf: resolve authorized input manifests once per execution

* test: preserve fetch signature in revocation fixture

* fix: isolate shared-grant failures and prevent ledger replay
* fix: evict stale file-object index entries

Forget cached locators after successful deletion and missing-object downloads so replacement keys resolve immediately. Reuse exact resolver matching in the delete route to avoid prefix collisions.

Fixes LibreChat-AI#181

* fix: keep file-object deletion storage-authoritative

* fix: retire superseded upload objects

* fix: canonicalize replacement object keys

* fix: collapse legacy object-key siblings

* fix: recover reads from stale locators

* fix: namespace canonical object identities
* fix: helm egress deployment getting stuck on install

* only wait for redis if the ledger is required

* Update helm/codeapi/templates/egress-gateway-deployment.yaml

Co-authored-by: Danny Avila <danacordially@gmail.com>

---------

Co-authored-by: Danny Avila <danacordially@gmail.com>
* 🛰️ feat: Add trusted VM command policy

* docs: clarify trusted VM socket boundary
…Chat-AI#188)

A clean atomic workspace mutation rejection was settled with retries cut off at the original execution deadline. When Stop arrived near that deadline, process-tree termination finished after it, so the first settlement attempt was aborted immediately while Code API was still draining the cancellation. The client received ASSIGNMENT_EXPIRED and the durable mutation guard stayed armed.

Route clean workspace mutation rejections through the known-clean rejection recovery path: a transient-retrying heartbeat and settlement retries through the rejection acknowledgement grace, floored at the bridge cancellation settlement grace. Worker shutdown still fails closed. Share the grace constant from the protocol module so the bridge and worker stay aligned.

Closes LibreChat-AI#173
…Chat-AI#191)

* fix: Exit Cleanly When Native Executor Shuts Down Concurrently

A native BYOM worker under systemd KillMode=control-group receives SIGTERM
at the same time as its forked SRT executor. The child ignores IPC once it
is shutting down, so the parent's close handshake is left pending until the
child exits, which rejects it with 'Native executor is unavailable'. That
rejection escaped the CLI finally block and turned an idle administrative
stop into exit status 1.

Treat the close handshake as best-effort: the executor is terminated in
finally regardless, and the active command has already drained, so a lost
or stalled reply carries no mutation risk. Also make the child report exit
status 0 when its own SRT teardown succeeded.

Closes LibreChat-AI#190

* fix: Surface Explicit Executor Cleanup Failures During Close

Only a lost, refused, or stalled close handshake is benign at shutdown.
A negative close reply from the executor is a real cleanup failure and
still rejects so pool shutdown can aggregate it.
* feat: run PTC in selected BYOM workspaces

* fix: harden native workspace PTC replay

* fix: preserve replay isolation and bridge limits

* test: tolerate hosts without filesystem cloning

* test: surface copy-on-write clone faults

* fix: harden native workspace PTC admission

* fix: close native replay effect and finalization boundaries
* feat: report truncated output artifacts

* fix: classify omitted artifacts precisely

* fix: preserve artifact scan invariants

* fix: bound depth truncation probes

* fix: bound capped directory enumeration

* fix: constrain truncation probes across the job

* fix: stop exhausted artifact probes
* fix: cancel replay jobs across API and worker processes

* fix: drain worker cancellation watches promptly

* fix: close programmatic cancellation races

* fix: preserve cancellation response ordering

* fix: close distributed cancellation races

* fix: harden cancellation under concurrent load

* fix: make cancellation ownership durable through completion

* fix: recover durable replay outcomes across lost replies

* fix: return atomic cancellation outcomes with aligned retention

* fix: commit native results inside the workspace mutation fence

* fix: claim programmatic execution before stalled-job redelivery
* fix: report deleted persisted files

* fix: reconcile deletions across code runtimes

* fix: preserve protected session inputs

* fix: classify reserved runtime paths
* feat: declare named worker project environments

* fix: preserve environment trust and negotiated action boundaries

* Harden environment loading and executor identity

* Protect environment root traversal and exact config bytes

* Reject self-controlled environment root aliases

* Check filesystem identities at environment trust boundaries

* Validate environment containment across Linux mount aliases

* Handle stacked mounts conservatively without blocking unrelated paths
* fix: Allow Trusted Own-Root Environment Symlinks

* fix: Check Alias Parent Ownership by Filesystem Identity

* fix: Enforce Parent Ownership Across Every Environment Path
* fix: Retain Quarantine After Failed Environment Setup

* test: Run Native Environment Setup Lifecycle in CI

* fix: Document and Verify Local Setup Quarantine Recovery
* fix: decouple repository release versions

* ci: automate releases after successful main builds
danny-avila and others added 13 commits September 16, 2026 01:51
* feat: add bounded local project inventory

* fix: report incomplete Git metadata reads

* fix: preserve incomplete discovery and remote identities

* fix: finalize discovery budgets and nested remote identities

* fix(code): stop project traversal at filesystem budget boundaries
LibreChat-AI#226)

* Discover bounded repository instructions for opted-in workspaces

* Verify snapshot digests and cross-platform confinement
* feat(code): register explicitly selected project roots

* fix(code): reject shared Git metadata for selected projects

* fix(code): pin selected project identity through executor admission

* fix(code): preserve full filesystem identity precision

* Check selected project identity before replay staging

* Anchor replay copies to the verified working directory

* fix: Bind Selected Project Operations to Held Directory Descriptors

* test: Cover Selected Project PTC and Load Native Fixtures Before Platform Simulation

* fix: Keep Native Root Bindings Worker-Local and Verify Directory Ancestry

* fix: Anchor Project Admission and Preserve Search Permissions
…rors (LibreChat-AI#227)

* fix: Distinguish workspace admission capacity from execution expiry

* test: Preserve execution uncertainty while classifying blocked follow-ups

* fix: Classify admission expiry at the enqueue boundary
…reChat-AI#233)

* ci: Fix Release Version Resolution for Untagged and Resumed Runs

The release workflow resolved its version in one inline shell block under
`set -euo pipefail`, where two paths could not succeed.

Filtering tags through `grep` made a no-match fatal. On the ordinary untagged
tip of `main`, `git tag --points-at HEAD | grep -E '^v[0-9]+...'` exits 1, and
the step died before reaching its skip handling or `next-release-version.sh`, so
a deployable commit could not obtain a release version (LibreChat-AI#228). Selecting stable
tags now reads exit 1 as an empty answer while exit 2 and above still fail the
release, which also lets the missing-previous-tag case report its own error.

The rerun-resume path then rejected the tag it had itself chosen. With a stable
tag already pointing at `HEAD` and no release published, the version comes from
that tag, and the following existence check failed merely because the ref
existed (LibreChat-AI#229). It now compares the tag's commit against the release commit, so
only a tag on some other commit is a collision; `Create tag` already tolerates
a tag that exists.

The block moved into `.github/scripts/resolve-release-version.sh`, beside the
`next-release-version.sh` it calls, so `tests/release-version-resolution.sh` can
cover every path: automatic, resumed, skipped, dispatched, pushed-tag, and the
runs that must be refused, each against a throwaway repository with a stubbed
`gh`.

* fix: harden release resolver execution

---------

Co-authored-by: Lia <lia@librechat.ai>
Co-authored-by: Danny Avila <danny@librechat.ai>
* feat: Route GitHub App credentials per repository

* test: Make repository routing assertion deterministic

* fix: Harden repository credential routing

* fix: Bound shared GitHub credential refreshes

* fix: Bind GitHub credentials to admitted workspaces
* feat: provision conversation-scoped code worktrees

* fix: isolate conversation checkout metadata

* docs: clarify isolated conversation checkouts

* fix: revalidate conversation checkout sources

* fix: preserve synchronous legacy execution startup

* fix: harden conversation worktree lifecycle

* test: use canonical workspace isolation keys

* fix: secure conversation worktree provisioning

* fix: preserve isolated workspace lifecycle

* fix: harden conversation worktree provisioning

* fix: fence worktree setup and credential routing

* fix: use kernel-backed provisioning locks

* fix: load worktree locking only when provisioned

* fix: retain conversation provisioning ownership through recovery

* fix: reserve provisioning before launching checkout writers

* fix: pin Git provisioning inputs and close instance admission gaps
…-AI#248)

* feat(code): route trusted VM GitHub App tokens by checkout

* fix(code): resolve checkout credentials within canonical root

* fix(code): reject cross-root credential aliases
@TomasPalsson
TomasPalsson merged commit e0d62ed into fix/nsjail-no-cgroup-clone Sep 25, 2026
10 checks passed
@TomasPalsson
TomasPalsson deleted the chore/sync-upstream-main branch September 25, 2026 13:06
TomasPalsson added a commit that referenced this pull request Sep 25, 2026
The upstream sync (#4) turned the minio import into a type-only import,
but the fork's createMinioClient still calls new Client(...). Bun strips
type imports, so Client is undefined at runtime and file-server exits with
'Critical: Could not initialize storage' on every start (seen on apro-sandbox,
task def code-interpreter:25). tsc reports it as TS1361 at lines 90/115/122.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants