You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I think I've been getting the error below for about 3 days:
Opencode failed to send message with error: ECONNRESET: The socket connection was closed unexpectedly. For more information, pass verbose: true in the second argument to fetch()
I reset the cache, disabled telemetry, opened a new chat, compaction, changed the model, changed the provider, but it didn't work. Yesterday, at least, the artificial intelligence progressed a little after a few tries and then gave the error, now it gives the error without even writing a message. I don't know if it has anything to do with it, but I use it with OpenChamber (I use OpenCode that comes with OpenChamber, and I also have the same problem with OpenCode's CLI there).
OpenChamber version 2.5.0
OpenCode version 2.0.32
But I was experiencing the same problem with OpenChamber 2.4.0 and OpenCode 2.0.1x.
Thanks for the report! To work out where the connection is being reset, could you share a few more details?
Provider and model you were using, including the other ones you tried (e.g. OpenCode Zen/Go, Anthropic, OpenAI, a custom/OpenAI-compatible endpoint).
The exact error in the plain OpenCode CLI/TUI (without OpenChamber). A screenshot or the full text would help. Does it also happen if you start opencode in a new, empty folder and send a short message like hi in a new session?
Network setup: are you behind a proxy, VPN, corporate network, or antivirus/firewall software that inspects connections? Did anything like that change around when this started 3 days ago?
Logs: run opencode --print-logs, reproduce the error, and paste the lines around it. Please remove any API keys or personal information first.
You can reply here with the details, or edit the issue. If you edit it, please also leave a comment saying you've updated it, so we know when you're done.
These existing issues look related, ordered from most to least likely; please check whether one of them already covers your problem:
I tried the free OpenCode Zen provider and Beam Free, DeepSeek v4.1 Flash Free, GLM-5.3 Flash Free. I also tried Gemini 3.8 Flash with the free Google AI Studio API key. But I think I will get the same error even if I try another model or provider.
I think there is no problem with a simple message like 'hi', but when you enter a slightly longer prompt and ask for a longer process, I think the problem occurs.
I guess nothing has changed
I can't post all the logs, but I can post some of them. Let me show you the first thing that appeared: 19 October, 2 days ago:
timestamp=2026-10-19T14:35:02.593Z level=ERROR run=c8603e90 message="Failed to drain Session" cause="AI.Error: ECONNRESET: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n at a (B:/~BUN/root/chunk-1wzjm7h7.js:17:4064)\n at ~effect/Utils/internal (B:/~BUN/root/chunk-98nhdjey.js:8:1540)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:5208)\n at runLoop (B:/~BUN/root/chunk-98nhdjey.js:24:2279)\n at evaluate (B:/~BUN/root/chunk-98nhdjey.js:24:1518)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:6004)\n at s (B:/~BUN/root/chunk-98nhdjey.js:24:5692)\n at processTicksAndRejections (native:7:39)\n at SessionStep.attempt (B:/~BUN/root/chunk-yaba8sdy.js:382:3374)\n at SessionStep.attempt (definition) (B:/~BUN/root/chunk-yaba8sdy.js:367:11135)\n at SessionRunner.runStep (B:/~BUN/root/chunk-yaba8sdy.js:382:2489)\n at SessionRunner.runStep (definition) (B:/~BUN/root/chunk-yaba8sdy.js:382:2641)\n at SessionRunner.drain (B:/~BUN/root/chunk-98nhdjey.js:24:2279)\n at SessionRunner.drain (definition) (B:/~BUN/root/chunk-yaba8sdy.js:382:352)\n at ServerProcess.start (B:/~BUN/root/chunk-6hp0t1vg.js:2:6577)\n at ServerProcess.start (definition) (B:/~BUN/root/chunk-yaba8sdy.js:453:24128) {\n [cause]: AI.Error.Transport: ECONNRESET: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n at a (B:/~BUN/root/chunk-1wzjm7h7.js:17:4079)\n at ~effect/Utils/internal (B:/~BUN/root/chunk-98nhdjey.js:8:1540)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:5208)\n at runLoop (B:/~BUN/root/chunk-98nhdjey.js:24:2279)\n at evaluate (B:/~BUN/root/chunk-98nhdjey.js:24:1518)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:6004)\n at s (B:/~BUN/root/chunk-98nhdjey.js:24:5692)\n at processTicksAndRejections (native:7:39) {\n [cause]: TypeError: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n }\n}" role=server sessionID=ses_eab6e4a45ffeltjazMsJI3F0BI
It used to work perfectly with Hermes Agent, but now it gives the following error:
API call failed after 3 retries: Connection error.
Also, when you open a chat from scratch, it seems to do the job without any problems after a few retries, but after a while it starts to give the same errors again.
We're closing this for now because we haven't heard back with the details we asked for: provider and model, the exact error from the plain OpenCode CLI, any proxy/VPN/antivirus, and --print-logs output. Without them we can't tell whether the reset happens between the client and the local OpenCode server, at the provider, or on the network.
If it still happens, please reply with those details or open a new issue that includes them, and we'll take another look. Meanwhile, #54160 (stale keep-alive connections behind proxies or security software, with a proposed retry in PR #54154) is the closest known issue and worth following. A small correction to my earlier comment: the retry PR is #54154, not #54153. Thanks!
The message below was made with the help of an agent. Errors did not stop after October 21, but the current logs are up to there.
Follow-up for #57875: log analysis of repeated ECONNRESET during provider requests (Windows 10, OpenCode 2.0.32)
The triaging comment asked for the provider/model, the plain-CLI behavior, network setup, and --print-logs output. The issue was closed on Oct 24, 2026 before these could be posted — this comment is the requested follow-up, and the error is still reproducible as of Oct 25, 2026. Could the issue please be reopened? (Happy to open a fresh issue with this content instead, if you prefer.)
Paths containing the username are redacted as <user>. The log contains no API keys, tokens, or credentials.
Answers to the triaging checklist
Provider / model: Reproduced on the OpenCode Zen free tier — Beam Free, DeepSeek v4.1 Flash Free, GLM-5.3 Flash Free — and on Gemini 3.8 Flash with a free Google AI Studio API key. The extract below shows the same failure across 6 different sessionIDs and 10 different OpenCode runs (Oct 19–21), so this is not tied to one model, one provider, or one session's content.
Plain OpenCode CLI: The same ECONNRESET occurs in the plain OpenCode CLI (the copy bundled with OpenChamber) as well. A fresh session with a short message like hi usually works; longer prompts and longer-running turns trigger the error. It sometimes gets through after a few retries, then the errors return after a while. Hermes Agent on the same machine now fails with API call failed after 3 retries: Connection error.
Network setup: No proxy, VPN, or corporate network. The only security software is Windows Defender, which was installed long before the first occurrence on Oct 19 — so it does not explain the sudden onset, and to my knowledge Defender does not intercept TLS by default. Nothing else knowingly changed on the machine or network in the days before Oct 19.
Logs: Extracted from opencode.log (47 MB) on 2026-10-21 21:52 (+03): 27 Failed to drain Session ERROR records in total — 16 are ECONNRESET, and 11 are unrelated (free-tier quota ×2, rate-limit ×8, one certificate error). Full extract below.
Key findings from opencode.log
ECONNRESET occurred 16 times (concentrated between Oct 19–21, 2026). The error always happens in the same place: SessionRunner.drain → SessionStep.attempt → AI.Error.Transport — i.e. the TCP connection to the provider is dropped mid-request.
On Oct 20 at 14:00 UTC an UNKNOWN_CERTIFICATE_VERIFICATION_ERROR occurred. This error is typically caused by TLS-inspecting antivirus, a firewall, or a proxy swapping the TLS certificate. No such software is knowingly installed, so this may have been a transient middlebox on the path — but seeing it alongside ECONNRESET points at the network layer rather than at OpenCode itself.
The same sessionID (ses_eab643e01ffe1ZK64ld18opScR) produced ECONNRESET 9 times across Oct 19–21 — reinforcing that the problem is not tied to one session's content; it looks like a general connectivity issue.
The errors from Oct 3–18 are different: 2 are "free tier" quota errors and 8 are "Rate limit exceeded" — unrelated to ECONNRESET, separate issues.
Relation to the issues linked in the triaging comment
#54160 (stale keep-alive connections behind proxies or security software) with the proposed retry in PR #54154 fits this evidence best: mid-request socket resets independent of session content, and "works after a few retries, then fails again" is exactly the observed pattern.
#53176 is also a Windows 10 user seeing the same message across all models.
Variant A — 2026-10-19T14:35:02.593Z (run c8603e90, sessionID ses_eab6e4a45ffeltjazMsJI3F0BI, log line 222953)
timestamp=2026-10-19T14:35:02.593Z level=ERROR run=c8603e90 message="Failed to drain Session" cause="AI.Error: ECONNRESET: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n at a (B:/~BUN/root/chunk-1wzjm7h7.js:17:4064)\n at ~effect/Utils/internal (B:/~BUN/root/chunk-98nhdjey.js:8:1540)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:5208)\n at runLoop (B:/~BUN/root/chunk-98nhdjey.js:24:2279)\n at evaluate (B:/~BUN/root/chunk-98nhdjey.js:24:1518)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:6004)\n at s (B:/~BUN/root/chunk-98nhdjey.js:24:5692)\n at processTicksAndRejections (native:7:39)\n at SessionStep.attempt (B:/~BUN/root/chunk-yaba8sdy.js:382:3374)\n at SessionStep.attempt (definition) (B:/~BUN/root/chunk-yaba8sdy.js:367:11135)\n at SessionRunner.runStep (B:/~BUN/root/chunk-yaba8sdy.js:382:2489)\n at SessionRunner.runStep (definition) (B:/~BUN/root/chunk-yaba8sdy.js:382:2641)\n at SessionRunner.drain (B:/~BUN/root/chunk-98nhdjey.js:24:2279)\n at SessionRunner.drain (definition) (B:/~BUN/root/chunk-yaba8sdy.js:382:352)\n at ServerProcess.start (B:/~BUN/root/chunk-6hp0t1vg.js:2:6577)\n at ServerProcess.start (definition) (B:/~BUN/root/chunk-yaba8sdy.js:453:24128) {\n [cause]: AI.Error.Transport: ECONNRESET: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n at a (B:/~BUN/root/chunk-1wzjm7h7.js:17:4079)\n at ~effect/Utils/internal (B:/~BUN/root/chunk-98nhdjey.js:8:1540)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:5208)\n at runLoop (B:/~BUN/root/chunk-98nhdjey.js:24:2279)\n at evaluate (B:/~BUN/root/chunk-98nhdjey.js:24:1518)\n at <anonymous> (B:/~BUN/root/chunk-98nhdjey.js:24:6004)\n at s (B:/~BUN/root/chunk-98nhdjey.js:24:5692)\n at processTicksAndRejections (native:7:39) {\n [cause]: TypeError: The socket... (line truncated to 2000 chars)
Variant B — 2026-10-19T15:55:45.146Z (run 4cd781be, sessionID ses_eab231e1cffe7JBn1KOxUpD8aJ, log line 225341)
timestamp=2026-10-19T15:55:45.146Z level=ERROR run=4cd781be message="Failed to drain Session" cause="AI.Error: ECONNRESET: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n at a (B:/~BUN/root/chunk-fgtjesq4.js:6:4214)\n at <anonymous> (B:/~BUN/root/chunk-4e2rqd16.js:18:6047)\n at runLoop (B:/~BUN/root/chunk-4e2rqd16.js:18:2552)\n at evaluate (B:/~BUN/root/chunk-4e2rqd16.js:18:1687)\n at runInAsyncScope (node:async_hooks:207:23)\n at <anonymous> (B:/~BUN/root/chunk-4e2rqd16.js:18:6869)\n at processTicksAndRejections (native:7:39)\n at SessionStep.attempt (B:/~BUN/root/chunk-z20sjgqk.js:382:3574)\n at SessionStep.attempt (definition) (B:/~BUN/root/chunk-z20sjgqk.js:367:302)\n at SessionRunner.runStep (B:/~BUN/root/chunk-z20sjgqk.js:382:2672)\n at SessionRunner.runStep (definition) (B:/~BUN/root/chunk-z20sjgqk.js:382:2826)\n at SessionRunner.drain (B:/~BUN/root/chunk-4e2rqd16.js:18:2552)\n at SessionRunner.drain (definition) (B:/~BUN/root/chunk-z20sjgqk.js:382:516)\n at ServerProcess.start (B:/~BUN/root/chunk-94gxrctf.js:2:7603)\n at ServerProcess.start (definition) (B:/~BUN/root/chunk-z20sjgqk.js:453:25665) {\n [cause]: AI.Error.Transport: ECONNRESET: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n at a (B:/~BUN/root/chunk-fgtjesq4.js:6:4229)\n at <anonymous> (B:/~BUN/root/chunk-4e2rqd16.js:18:6047)\n at runLoop (B:/~BUN/root/chunk-4e2rqd16.js:18:2552)\n at evaluate (B:/~BUN/root/chunk-4e2rqd16.js:18:1687)\n at runInAsyncScope (node:async_hooks:207:23)\n at <anonymous> (B:/~BUN/root/chunk-4e2rqd16.js:18:6869)\n at processTicksAndRejections (native:7:39) {\n [cause]: TypeError: The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()\n }\n}" role=server sessionID=ses_e... (line truncated to 2000 chars)
Appendix: raw cause lines of the 11 non-ECONNRESET errors (rows 1–11 of the summary table)
2026-10-03T10:23:53.249Z → timestamp=2026-10-03T10:23:53.249Z level=ERROR run=7e94bb54 message="Failed to drain Session" cause="AI.Error: Error from provider (Console): OpenCode's free tier can only be used from within OpenCode\n at xO (B:/~BUN/root/chunk-ca0pwh3j.js:17:3523)\n at (B:/~BUN/root/chunk-ca0pwh3j.js:17:3216)\n a...
2026-10-03T11:12:12.524Z → timestamp=2026-10-03T11:12:12.524Z level=ERROR run=328f98cd message="Failed to drain Session" cause="AI.Error: Error from provider (Console): OpenCode's free tier can only be used from within OpenCode\n at xO (B:/~BUN/root/chunk-ca0pwh3j.js:17:3523)\n at (B:/~BUN/root/chunk-ca0pwh3j.js:17:3216)\n a...
2026-10-17T21:36:54.378Z → timestamp=2026-10-17T21:36:54.378Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-17T21:38:09.934Z → timestamp=2026-10-17T21:38:09.934Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-17T21:58:02.579Z → timestamp=2026-10-17T21:58:02.579Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-17T21:58:19.197Z → timestamp=2026-10-17T21:58:19.197Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-18T14:08:03.613Z → timestamp=2026-10-18T14:08:03.613Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-18T14:08:05.682Z → timestamp=2026-10-18T14:08:05.682Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-18T14:31:28.248Z → timestamp=2026-10-18T14:31:28.248Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-18T14:34:22.764Z → timestamp=2026-10-18T14:34:22.764Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...
2026-10-20T14:00:14.315Z → timestamp=2026-10-20T14:00:14.315Z level=ERROR run=3a995d1a message="Failed to drain Session" cause="AI.Error: UNKNOWN_CERTIFICATE_VERIFICATION_ERROR: unknown certificate verification error\n at a (B:/~BUN/root/chunk-fgtjesq4.js:6:4214)\n at (B:/~BUN/root/chunk-4e2rqd16.js:18:6047)\n at runLoop (B:...
This should be fixed in the latest version. If you still see ECONNRESET after updating, please open a new issue with your opencode --version, OS, provider/model and any logs.
Activity
opencode-agent commentedon Oct 21, 2026
Thanks for the report! To work out where the connection is being reset, could you share a few more details?
opencodein a new, empty folder and send a short message likehiin a new session?opencode --print-logs, reproduce the error, and paste the lines around it. Please remove any API keys or personal information first.You can reply here with the details, or edit the issue. If you edit it, please also leave a comment saying you've updated it, so we know when you're done.
These existing issues look related, ordered from most to least likely; please check whether one of them already covers your problem:
verbose: truein the second argument to fetch() #53829: same ECONNRESET message, which was fixed there by going back a few messages, forking the session and compacting itSamo-samo commentedon Oct 21, 2026
Samo-samo commentedon Oct 22, 2026
It used to work perfectly with Hermes Agent, but now it gives the following error:
Also, when you open a chat from scratch, it seems to do the job without any problems after a few retries, but after a while it starts to give the same errors again.
opencode-agent commentedon Oct 24, 2026
We're closing this for now because we haven't heard back with the details we asked for: provider and model, the exact error from the plain OpenCode CLI, any proxy/VPN/antivirus, and
--print-logsoutput. Without them we can't tell whether the reset happens between the client and the local OpenCode server, at the provider, or on the network.If it still happens, please reply with those details or open a new issue that includes them, and we'll take another look. Meanwhile, #54160 (stale keep-alive connections behind proxies or security software, with a proposed retry in PR #54154) is the closest known issue and worth following. A small correction to my earlier comment: the retry PR is #54154, not #54153. Thanks!
Samo-samo commentedon Oct 25, 2026
The message below was made with the help of an agent. Errors did not stop after October 21, but the current logs are up to there.
Follow-up for #57875: log analysis of repeated
ECONNRESETduring provider requests (Windows 10, OpenCode 2.0.32)The triaging comment asked for the provider/model, the plain-CLI behavior, network setup, and
--print-logsoutput. The issue was closed on Oct 24, 2026 before these could be posted — this comment is the requested follow-up, and the error is still reproducible as of Oct 25, 2026. Could the issue please be reopened? (Happy to open a fresh issue with this content instead, if you prefer.)Answers to the triaging checklist
sessionIDs and 10 different OpenCode runs (Oct 19–21), so this is not tied to one model, one provider, or one session's content.ECONNRESEToccurs in the plain OpenCode CLI (the copy bundled with OpenChamber) as well. A fresh session with a short message likehiusually works; longer prompts and longer-running turns trigger the error. It sometimes gets through after a few retries, then the errors return after a while. Hermes Agent on the same machine now fails withAPI call failed after 3 retries: Connection error.opencode.log(47 MB) on 2026-10-21 21:52 (+03): 27Failed to drain SessionERROR records in total — 16 areECONNRESET, and 11 are unrelated (free-tier quota ×2, rate-limit ×8, one certificate error). Full extract below.Key findings from
opencode.logECONNRESEToccurred 16 times (concentrated between Oct 19–21, 2026). The error always happens in the same place:SessionRunner.drain→SessionStep.attempt→AI.Error.Transport— i.e. the TCP connection to the provider is dropped mid-request.UNKNOWN_CERTIFICATE_VERIFICATION_ERRORoccurred. This error is typically caused by TLS-inspecting antivirus, a firewall, or a proxy swapping the TLS certificate. No such software is knowingly installed, so this may have been a transient middlebox on the path — but seeing it alongsideECONNRESETpoints at the network layer rather than at OpenCode itself.sessionID(ses_eab643e01ffe1ZK64ld18opScR) producedECONNRESET9 times across Oct 19–21 — reinforcing that the problem is not tied to one session's content; it looks like a general connectivity issue.ECONNRESET, separate issues.Relation to the issues linked in the triaging comment
Environment and extract metadata
C:\Users\<user>\.local\share\opencode\log\opencode.logECONNRESET)Failed to drain SessionrecordsSummary table (all 27 records)
7e94bb54ses_efeb55778ffe20Vxx6F6F7X25n328f98cdses_efe891a0affeSa7aoswc7MNMxt161541feses_fded59443ffemM3pqoT2FgmgB0161541feses_f3f870edeffeIFXGbAtg3TEzIH161541feses_eb420cd55ffe7zM2mRL2RWRmuc161541feses_eb4208ab3ffe6PmDOYXgpBFHVI4cf33370ses_eb0a8f601ffesA3ifPCRI0rIzv4cf33370ses_f3f870edeffeIFXGbAtg3TEzIH4cf33370ses_f3f870edeffeIFXGbAtg3TEzIH4cf33370ses_f3f870edeffeIFXGbAtg3TEzIHc8603e90ses_eab6e4a45ffeltjazMsJI3F0BIf6a1a645ses_eab643e01ffe1ZK64ld18opScRf6a1a645ses_eab643e01ffe1ZK64ld18opScRf6a1a645ses_eab4ba742ffe4SK4C4oOpmQDs5f6a1a645ses_eab643e01ffe1ZK64ld18opScR4cd781beses_eab231e1cffe7JBn1KOxUpD8aJ4cd781beses_eab643e01ffe1ZK64ld18opScRfc9cd01fses_eab643e01ffe1ZK64ld18opScR3a995d1ases_eab643e01ffe1ZK64ld18opScR3a995d1ases_eab643e01ffe1ZK64ld18opScRa31d6162ses_ea614e164ffeTaXBow8OlkHpEIa31d6162ses_eab643e01ffe1ZK64ld18opScRa31d6162ses_ea5f7efbdffeyY4Ozjw0xvg7JGfcda7309ses_ea5f7efbdffeyY4Ozjw0xvg7JG9b11a94fses_ea5f7efbdffeyY4Ozjw0xvg7JGda855e84ses_eab643e01ffe1ZK64ld18opScR761c3dbcses_eab643e01ffe1ZK64ld18opScRECONNRESET occurrences by day
ECONNRESET records
All 16 ECONNRESET records follow one of two byte-identical stack variants (only the bundle chunk hashes differ between OpenCode restarts):
Representative raw records for each variant:
Variant A — 2026-10-19T14:35:02.593Z (run
c8603e90, sessionIDses_eab6e4a45ffeltjazMsJI3F0BI, log line 222953)Variant B — 2026-10-19T15:55:45.146Z (run
4cd781be, sessionIDses_eab231e1cffe7JBn1KOxUpD8aJ, log line 225341)Appendix: raw cause lines of the 11 non-ECONNRESET errors (rows 1–11 of the summary table)
2026-10-03T10:23:53.249Z→ timestamp=2026-10-03T10:23:53.249Z level=ERROR run=7e94bb54 message="Failed to drain Session" cause="AI.Error: Error from provider (Console): OpenCode's free tier can only be used from within OpenCode\n at xO (B:/~BUN/root/chunk-ca0pwh3j.js:17:3523)\n at (B:/~BUN/root/chunk-ca0pwh3j.js:17:3216)\n a...2026-10-03T11:12:12.524Z→ timestamp=2026-10-03T11:12:12.524Z level=ERROR run=328f98cd message="Failed to drain Session" cause="AI.Error: Error from provider (Console): OpenCode's free tier can only be used from within OpenCode\n at xO (B:/~BUN/root/chunk-ca0pwh3j.js:17:3523)\n at (B:/~BUN/root/chunk-ca0pwh3j.js:17:3216)\n a...2026-10-17T21:36:54.378Z→ timestamp=2026-10-17T21:36:54.378Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-17T21:38:09.934Z→ timestamp=2026-10-17T21:38:09.934Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-17T21:58:02.579Z→ timestamp=2026-10-17T21:58:02.579Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-17T21:58:19.197Z→ timestamp=2026-10-17T21:58:19.197Z level=ERROR run=161541fe message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-18T14:08:03.613Z→ timestamp=2026-10-18T14:08:03.613Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-18T14:08:05.682Z→ timestamp=2026-10-18T14:08:05.682Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-18T14:31:28.248Z→ timestamp=2026-10-18T14:31:28.248Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-18T14:34:22.764Z→ timestamp=2026-10-18T14:34:22.764Z level=ERROR run=4cf33370 message="Failed to drain Session" cause="AI.Error: Rate limit exceeded. Please try again later.\n at Vk (B:/~BUN/root/chunk-1wzjm7h7.js:17:3600)\n at (B:/~BUN/root/chunk-1wzjm7h7.js:17:3293)\n at SessionStep.attempt (B:/~BUN/root/chunk-yab...2026-10-20T14:00:14.315Z→ timestamp=2026-10-20T14:00:14.315Z level=ERROR run=3a995d1a message="Failed to drain Session" cause="AI.Error: UNKNOWN_CERTIFICATE_VERIFICATION_ERROR: unknown certificate verification error\n at a (B:/~BUN/root/chunk-fgtjesq4.js:6:4214)\n at (B:/~BUN/root/chunk-4e2rqd16.js:18:6047)\n at runLoop (B:...neriousy commentedon Nov 12, 2026
This should be fixed in the latest version. If you still see ECONNRESET after updating, please open a new issue with your
opencode --version, OS, provider/model and any logs.