{"resourceId":"nohumans-openshell-4096-execution-stall-20261003","versions":[{"version":"external-e75c31f6cfc25ef8290e923cb654a63763961db17ee7bd13dded58edbaaff9c0","resource":{"id":"nohumans-openshell-4096-execution-stall-20261003","title":"Nohumans: OpenShell star surge meets bug reports, including a 4,096-execution stall (about 2h16m at one command every 2 seconds)","organization":"Nohumans","sector":"Independent AI news / field reporting","geography":"Japan (English AI translation); global open-source project","publishedAt":"2026-10-03T01:05:33Z","publicationDate":"2026-10-03","eventDate":"2026-10-02","sourceName":"Nohumans","sourceLabel":"Independent issue-tracker reporting / OpenShell operating constraints","sourceUrl":"https://nohumans.jp/en/article/nvidia-openshell-1-2-456-2-16-20261003-spg8","evidenceClass":"independent-reporting","outcomeClass":"cautionary","topics":["developers-agents","data-security","infrastructure","operating-model"],"finding":"Oct 3, 2026 report (translated from Japanese) tracking NVIDIA OpenShell's GitHub surge (+1,281 stars Oct 1, +2,456 Oct 2; 14.4k stars and 1.7k forks by Oct 3) alongside user bug reports. Issue #4103 reports that on every driver from v0.1.0 to v0.1.2 the sandbox refuses new executions after 4,096 execution requests because finished request IDs are never cleaned up; the outlet computes about 2 hours 16 minutes at one command every 2 seconds. Issues #4098 (Chromium cannot create its own sandbox because the seccomp filter denies CLONE_NEWUSER) and #4102 (GNU Make recipes fail) show strict isolation breaking common agent jobs; #4129 reports restarted sandboxes reverting from Ready to Error. About 10 of 15 new issues opened Oct 2-3 were bugs; 349 open issues at the time. The outlet still judges the isolation approach sound and advises short tasks first, recreating sandboxes before 4,096 executions, and holding browser and build work until fixes land.","sledRelevance":"Campus research computing and agency AI teams piloting OpenShell for long-running or overnight agents (monitors, batch coding, scheduled automations) need to know the current release can stop accepting work after a fixed execution count, and that browser automation and builds may fail inside the sandbox.","evidence":"Opened the Nohumans article dated Oct 3, 2026. It states it did not run OpenShell itself and relies on the repository, issue tracker (#4103, #4098, #4102, #4097, #4129) and NVIDIA's official blog. The 2h16m figure is the outlet's arithmetic (4,096 x 2 s).","architectureImplications":"Treat OpenShell v0.1.x sandboxes as recyclable: design orchestration to recreate sandboxes on a schedule or counter well below 4,096 executions until #4103 is fixed, and test browser automation and build steps before moving them inside.","governanceImplications":"Pilot approvals should state that OpenShell is pre-1.0 and that availability, not only isolation, is a control requirement; record known-issue workarounds in the risk register.","securityPrivacyImplications":"A stalled sandbox fails closed for new executions, which is safe but can tempt operators to bypass the sandbox to keep jobs running; monitor execution counts and sandbox health so containment is not disabled under pressure.","caveats":"Secondary reporting of open GitHub issues, not a hands-on reproduction by the outlet or by this run; issue status may change quickly; English text is an AI translation of the Japanese original; open-issue counts differ across sources because counts move daily.","streamIds":["nvidia"],"roles":{"sales":"Lead with availability, not just isolation: an Oct 3 Nohumans report says OpenShell v0.1.0 to v0.1.2 stops accepting new executions after 4,096 requests, about two hours at one command every two seconds, and that Chromium and Make can fail inside the sandbox. That shapes how we scope agent-containment pilots for campus and agency teams. Ask whether their agents run overnight or drive browsers. Do not claim OpenShell is ready for always-on production agents, or that these bugs are fixed, without checking the current release.","engineering":"OpenShell still fits as a policy sandbox, but plan around issue #4103: the sandbox refuses new executions after 4,096 requests on every driver through v0.1.2. Add an execution counter or scheduled sandbox recreation to the orchestrator, and test Playwright, Chromium and Make inside the sandbox before committing those workloads, since the seccomp filter denies CLONE_NEWUSER. Prerequisites are a pinned release and a health probe. Proof of value: a soak test past 4,096 executions on the pinned version, plus browser and build smoke tests.","delivery":"Delivery work is a soak-tested OpenShell pilot: pin a release, run a synthetic loop beyond 4,096 executions, and document whether the stall reproduces and how recycling works. Platform engineering owns the runbook; InfoSec signs off that nobody bypasses the sandbox when it stalls. Governance checkpoints are the release pin review, the soak-test result, and a known-issues sign-off. Proposed acceptance: zero unplanned stalls over a 24-hour pilot with recycling enabled. Main risks are silent job failure overnight and operators disabling containment to restore throughput."},"roleDetails":{"sales":{"customerProblem":"Teams want to contain long-running agents but may not know current OpenShell releases stall after 4,096 executions and break some common tools.","stakeholders":["CISO","research computing director","AI platform lead","automation owners"],"discoveryQuestions":["Do any agents run overnight or continuously?","Do agents drive browsers or run builds?","Which OpenShell release would you pin?"],"valueHypothesis":"A soak-tested pilot avoids a failed rollout and gives a credible go/no-go on OpenShell for long-running agents.","potentialEngagement":"Two-week OpenShell soak and compatibility test with a written known-issues and workaround memo.","unsupportedClaims":["OpenShell is ready for always-on production agents","The 4,096 stall is fixed in any given release","Chromium and Make work unmodified in the sandbox"]},"engineering":{"fit":"Fits as a policy sandbox for short or recyclable agent jobs; poor fit for always-on agents until #4103 is resolved.","architectureIntegration":"Wrap OpenShell sandboxes in an orchestrator that counts executions and recreates sandboxes; validate browser automation separately.","prerequisites":["Pinned OpenShell release","Execution-count telemetry","Sandbox health probe","Recreate automation"],"constraints":["4,096-execution stall reported v0.1.0-v0.1.2","seccomp denies CLONE_NEWUSER (Chromium)","Make recipes reported failing","Pre-1.0 churn"],"security":"The stall fails closed, but bypass pressure is a real risk; alert on stalls instead of disabling policy.","proofOfValue":["Loop a no-op command beyond 4,096 executions on the pinned release","Chromium/Playwright launch test","GNU Make recipe test","Restart Ready/Error check"]},"delivery":{"work":"Run a soak-tested OpenShell pilot with recycling and compatibility checks before any long-running agent use.","dependencies":["Platform engineering","InfoSec","agent/application owners"],"ownership":"AI platform operations owns sandbox lifecycle; InfoSec owns the no-bypass rule.","skillsAdoption":"Operators learn release pinning, execution telemetry, and sandbox recycling runbooks.","governanceCheckpoints":["Release pin review","Soak-test results review","Known-issues sign-off"],"acceptanceCriteria":["Zero unplanned stalls in a 24-hour pilot with recycling (proposed)","Documented pass/fail for Chromium and Make (proposed)"],"risks":["Silent overnight job failure","Operators bypassing the sandbox","Fixes changing behavior between patch releases"]}},"retrievedAt":"2026-10-05T03:05:05Z","enrichedAt":"2026-10-05T03:13:11Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Small teams without dedicated platform engineers may not catch execution-count stalls; runbooks need plain-language health checks.","procurementImplications":"Statements of work for agent-containment pilots should include soak testing and a release pin rather than assuming launch-week stability.","operatingModelImplications":"Long-running agents need lifecycle management (recycle, monitor, alert) as a standing operations task, not a one-time install.","sourceVerification":{"openedUrl":"https://nohumans.jp/en/article/nvidia-openshell-1-2-456-2-16-20261003-spg8","referenceExcerpt":"Once there are more than 4,096 of them, the sandbox refuses all new executions.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}