anomalyco/opencode
1
0
Fork 0
mirror of https://github.com/anomalyco/opencode.git synced 2026-07-22 10:15:31 -04:00
Code Issues 5.9k Packages Projects Releases 19 Wiki Activity

Where does OpenCode expect us to receive updates about its recent API changes c.f. Big Pickle? #7464

New Issue
Open
opened 2026-02-16 18:07:16 -05:00 by yindo · 5 comments
yindo commented 2026-02-16 18:07:16 -05:00
Owner
Copy Link
Copy Source

Originally created by @davidbernat on GitHub (Jan 24, 2026).

Originally assigned to: @fwang on GitHub.

Description

OpenCode team-- beginning this morning Big Pickle has started throwing up massive throttle problems on every OpenCode prompt cycle. Since, as we previously requested, do not have transparency into the OpenCode prompt actions, we have know way of knowing how many Big Pickle API queries OpenCode is deciding (based on its own investor backed software) to sent to Big Pickle. It could be one (my prompt sent to Big Pickle via Zen) or it could be thousands. (After all, the Internet laughed when Elon Musk famously engineered Twitter, oh, whatever the microservices saga was.) So unfortunately OpenCode does not provide tools for this user (me) to diagnose this?

Presumably, OpenCode and Zen made the decision last night. Big Pickle had been malfunctioning all evening, with wrong answers and broken tool calls, and we made quite a large number of post-mortem documentation within our company (some even reached YC). Of course, using Big Pickle in which one first session prompt generates "more than 1 minute 17 seconds of wait" is unusable. So we kindly as how we were otherwise expected to be informed about this change at OpenCode? Was an email sent out that we are not subscribers to?

Yesterday, prior to these emails we send, Big Pickle went on an insane Arkham Asylum level year through our code repositories, deleting git caches, git commits, rm deleting entire swaths of files, and tons of session logs for reconstructing the day prior. We joked internally that my previous warnings on OpenCode were equivalent to firing a junior engineer for boneheaded never-make decisions live during a board room demonstration, but that these decisions yesterday were so unheard of in nominal code development that were an engineer to do this we would presume they were torching the company after their dismissal and be subject to criminal legal actions. (We are of course not asserting that criminal action is taken against OpenCode for a couple of missing session files. We simply run our operation with the thorough seriousness of a captain of a battle brigade and recognize that every client expects their company to be conducted with integrity or else OpenCode solutions are valueless to them. We surely reinstated our code and moved on with OUR life.)

So what is up with Big Pickle?
TLDR: How were we supposed to learn about this change of consumer relationship to Big Pickle API allocation?

What makes this especially harrowing for such relationships is that we had just begun [in our second internal project, operating in isolation] testing OpenCode for the real moneymakers of your service, i.e., system administration trust at the Linux command line. As I think I said previously, my resume is available online if you boys want to hire me to help engineer some of these solutions.

It is rare, to work with a master team. Thank you for letting me know. LinkedIn & YC are investors, so we know.
Image

Image
Originally created by @davidbernat on GitHub (Jan 24, 2026). Originally assigned to: @fwang on GitHub. ### Description OpenCode team-- beginning this morning Big Pickle has started throwing up massive throttle problems on every OpenCode prompt cycle. Since, as we previously requested, do not have transparency into the OpenCode prompt actions, we have know way of knowing how many Big Pickle API queries OpenCode is deciding (based on its own investor backed software) to sent to Big Pickle. It could be one (my prompt sent to Big Pickle via Zen) or it could be thousands. (After all, the Internet laughed when Elon Musk famously engineered Twitter, oh, whatever the microservices saga was.) So unfortunately OpenCode does not provide tools for this user (me) to diagnose this? Presumably, OpenCode and Zen made the decision last night. Big Pickle had been malfunctioning all evening, with wrong answers and broken tool calls, and we made quite a large number of post-mortem documentation within our company (some even reached YC). Of course, using Big Pickle in which one first session prompt generates "more than 1 minute 17 seconds of wait" is unusable. So we kindly as how we were otherwise expected to be informed about this change at OpenCode? Was an email sent out that we are not subscribers to? Yesterday, prior to these emails we send, Big Pickle went on an insane Arkham Asylum level year through our code repositories, deleting git caches, git commits, rm deleting entire swaths of files, and tons of session logs for reconstructing the day prior. We joked internally that my previous warnings on OpenCode were equivalent to firing a junior engineer for boneheaded never-make decisions live during a board room demonstration, but that these decisions yesterday were so unheard of in nominal code development that were an engineer to do this we would presume they were torching the company after their dismissal and be subject to criminal legal actions. (We are of course not asserting that criminal action is taken against OpenCode for a couple of missing session files. We simply run our operation with the thorough seriousness of a captain of a battle brigade and recognize that every client expects their company to be conducted with integrity or else OpenCode solutions are valueless to them. We surely reinstated our code and moved on with OUR life.) So what is up with Big Pickle? TLDR: How were we supposed to learn about this change of consumer relationship to Big Pickle API allocation? What makes this especially harrowing for such relationships is that we had just begun [in our second internal project, operating in isolation] testing OpenCode for the real moneymakers of your service, i.e., system administration trust at the Linux command line. As I think I said previously, my resume is available online if you boys want to hire me to help engineer some of these solutions. It is rare, to work with a master team. Thank you for letting me know. LinkedIn & YC are investors, so we know. <img width="760" height="130" alt="Image" src="https://github.com/user-attachments/assets/bd750092-5b57-40c1-ba9a-534bb9fdcc0b" /> <img width="784" height="373" alt="Image" src="https://github.com/user-attachments/assets/28d020c4-1ec5-449d-8413-d2e74923dfcf" />
yindo added the bugperfzen labels 2026-02-16 18:07:16 -05:00
yindo commented 2026-02-16 18:07:17 -05:00
Author
Owner
Copy Link
Copy Source

@github-actions[bot] commented on GitHub (Jan 24, 2026):

This issue might be a duplicate of existing issues. Please check:

  • #10404: Big Pickle model loops with "too many requests" error in high mode (free trial)
  • #9248: Big Pickle via OpenCode Zen is down this morning?
  • #10088: Zen Upgrade: All models getting stuck/unable to complete sessions

Feel free to ignore if none of these address your specific case.

@github-actions[bot] commented on GitHub (Jan 24, 2026): This issue might be a duplicate of existing issues. Please check: - #10404: Big Pickle model loops with "too many requests" error in high mode (free trial) - #9248: Big Pickle via OpenCode Zen is down this morning? - #10088: Zen Upgrade: All models getting stuck/unable to complete sessions Feel free to ignore if none of these address your specific case.
yindo commented 2026-02-16 18:07:17 -05:00
Author
Owner
Copy Link
Copy Source

@rekram1-node commented on GitHub (Jan 24, 2026):

This ratelimiting is done via ip, we should prolly scale it up since it is getting hit harder recently

@rekram1-node commented on GitHub (Jan 24, 2026): This ratelimiting is done via ip, we should prolly scale it up since it is getting hit harder recently
yindo commented 2026-02-16 18:07:17 -05:00
Author
Owner
Copy Link
Copy Source

@fwang commented on GitHub (Jan 24, 2026):

@davidbernat we just pushed out a change.. do you still get rate limit errors?

@fwang commented on GitHub (Jan 24, 2026): @davidbernat we just pushed out a change.. do you still get rate limit errors?
yindo commented 2026-02-16 18:07:17 -05:00
Author
Owner
Copy Link
Copy Source

@brenai-works commented on GitHub (Jan 25, 2026):

Yeh, this change was new to me when i started using OpenCode today. I was using one of the other free models for the last month to test out the agent's UX, and just noticed the switch by default. There isn't much information out there about Big Pickle or the Zen in-house model which a little disappointing.

I think these sort of problems are going to be significant bottlenecks for open-source coding agents moving forward [...]

@brenai-works commented on GitHub (Jan 25, 2026): Yeh, this change was new to me when i started using OpenCode today. I was using one of the other free models for the last month to test out the agent's UX, and just noticed the switch by default. There isn't much information out there about Big Pickle or the Zen in-house model which a little disappointing. I think these sort of problems are going to be significant bottlenecks for open-source coding agents moving forward [...]
yindo commented 2026-02-16 18:07:17 -05:00
Author
Owner
Copy Link
Copy Source

@davidbernat commented on GitHub (Feb 12, 2026):

@fwang @rekram1-node Absolutely suggesting that I be able to access my IP meta-usage statistics at any point to see my per-day artifacts. Congratulations on being rate limited by your hardware; a true measure of scale and success. But, again, as I try to emphasize, as a business user in a boardroom I need to know when backend changes are happening with the served models.

To the point of @brenai-works we have no idea what Big Pickle is, who provides this, what the ELA/EUA is for our data, or even how large the GPUs are which carry its core, or whether lower level models are swapped out at cost reasons or capacity issues.

Can you imagine the horrors of me unleashing OpenCode v2.1.21 on a tightly controlled closed-loop proprietary daytona stack of wiggum loops, only to discover that someone at Big Pickle Rick Inc. swapped in o4-mini for o4-maxi because a compliance officer resigned in protest in front of the investors again? What would I be saying to my compliance officer at Starlight AutoInc.

You know how it is. Thanks.

@davidbernat commented on GitHub (Feb 12, 2026): @fwang @rekram1-node Absolutely suggesting that I be able to access my IP meta-usage statistics at any point to see my per-day artifacts. Congratulations on being rate limited by your hardware; a true measure of scale and success. But, again, as I try to emphasize, as a business user in a boardroom I need to know when backend changes are happening with the served models. To the point of @brenai-works we have no idea what Big Pickle is, who provides this, what the ELA/EUA is for our data, or even how large the GPUs are which carry its core, or whether lower level models are swapped out at cost reasons or capacity issues. Can you imagine the horrors of me unleashing OpenCode v2.1.21 on a tightly controlled closed-loop proprietary daytona stack of wiggum loops, only to discover that someone at Big Pickle Rick Inc. swapped in o4-mini for o4-maxi because a compliance officer resigned in protest in front of the investors again? What would I be saying to my compliance officer at Starlight AutoInc. You know how it is. Thanks.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
acp
bad
bug
discussion
docs
documentation
enhancement
extension
good first issue
growth
help-wanted
mcp
model-problem
needs:compliance
nix
opentui
perf
pull-request
Mirrored from GitHub Pull Request
 tui
web
windows
zen
No Label bug perf zen
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: anomalyco/opencode#7464
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?