All Issues
24,993 verified issues
Agent mode hangs when running terminal command
- Copilot Chat Extension Version: 0.26.2025040201 - VS Code Version: 1.99.0-insider (user setup) Commit: 8222926219175706ebc0af3c03db588d2bbe185a Date: 2025-04-01T22:33:55.946Z Electron: 34.3.2 ElectronBuildId: 11161073 Chromium: 132.0.6834.210 Node.js: 20.18.3 V8: 13.2.152.41-electron.0 OS: Windows_NT x64 10.0.22631 Steps to Reproduce: 1. prompt for any action that requires running a command in terminal, e.g. "build the project" I get the command to be executed shown and I click the button "Continue" A terminal opens where agent issues the command. A rotating "spinner" appears next to the command in chat window and that's it
Since docker 1.3.0 / fig 1.0.0 volumes are not mounted when fig up'ing existing containers
Since docker 1.3.0 / fig 1.0.0 (I upgraded both at the same time), I cannot fig up my projects without fig rm'ing before. If I do fig up on already existing (stopper) containers, volumes specified in fig.yml aren't remounted (default docker mounts instead, so empty dirs). fig rm && fig up recreates everything correctly but for locally built containers this is pretty time consuming.
Claude-2 100k model never outputs more than 4k tokens
Even though cloude-2 has 100k tokens context window I can not get it to generate more than 4k tokens. It interrupts generation after exactly 4096 tokens. Even if I set `max_tokens_to_sample` to be more than 4096. I tried both anthropic-sdk-python and web interface - both interrupt in the same place. Code to reproduce: [code block] If you run that code you can see, that completion interrupted mid-sentense after reaching 4096 tokens: [code block] It is the last chars of the output. And in developer log on anthropic page it looks like that: <img width="839" alt="Screenshot 2023-09-15 at 11 00 37" src="https://github.com/anthropics/anthropic-sdk-python/assets/120242470/6be42ad7-4038-41eb-9b4d-67c5d8aaa3d4"> So that 100k tokens only for input, and only 4k of them for output? It is very unexpected (because openai models not like that) and not documented behaviour.
Unknown Arguments
Hi! I have an ES6 class named 'Charge'. When I create a new instance of it and pass it to stripe.charges.create it returns an error (below). It does not when I pass a vanilla object. [code block] [code block]
[BUG] Passing CLI arguments via a Node scripts no longer works
Is there an existing issue for this? - [X] I have searched the existing issues This issue exists in the latest npm version - [X] I am using the latest npm Current Behavior With the following `index.js` NodeJS file in the root of a new clean NodeJS project: [code block] I get the following output when I run `node ./index.js -test-arg test-arg-value`: [code block] When I add an NPM script with: [code block] and then run it with either of the following: - `npm run arg-test -- -test-arg test-arg-value`, - `npm run arg-test -test-arg test-arg-value`, I simply get: [code block] This leaves me unable to run CLI tools, such as Yargs, via NPM scripts like we used to be able to. Expected Behavior When using NPM to compose reusable scripts, CLI arguments should still be supported - which enables the use of CLI tools such as Yargs Steps To Reproduce 1. Using the latest NodeJS (_20.12.0_) and the latest NPM (_10.5.2_) 2. With an `index.js` file in the root of a new NodeJS project containing the following: - `index.js`: [code block] 4. Run the file with `node ./index.js -test-arg test-arg-value` 5. See the expected output: `[ '-test-arg', 'test-arg-value' ]` 6. Add an NPM script to the `package.json` file to run the file: - `package.json`: [code block] 7. Run the file with either of the following: - `npm run arg-test -- -test-arg test-arg-value`, - `npm run arg-test -test-arg test-arg-value`, 8. See the incorrect output `[ 'test-arg-value' ]` Note
Sentinels set up with wrong master password not throwing when authentication fails
So not sure how to explain this quickly and cleanly but basically this is roughly the case scenario that happened due to some configuration not being (automatically) deployed correctly: 3 redis servers UP ips: 1, 2, 3 port 6379 server 1 is master (protected with password: password) server 2,3 slaves (correctly configured to be slave of server 1) 3 sentinels UP ips: 1,2,3 port 26379 Sentinels are badly configured /etc/sentinel.conf sentinel 1 monitor master 1 (correctly configured to monitor the master with the right password) sentinel 2 monitor master 2 (2 is not a master) <- bad configuration (master ip is that of the slave ip ) sentinel 3 monitor master 3 (3 is not a master) <- bad configuration (master ip is that of the slave ip ) with quorum set to 2 With this set up [code block] The connect event is being raised (Right we can connect to the sentinels) The Ready event is never being raised (Right the sentinels cannot communicate / agree / authenticate to the master ) However no error event being raised.... Hence we don't know from a client side perspective what it is wrong even though the connection to sentinels is established Sending set redis commands also results in no errors being raised and (nothing actually gets set in the cache) Does this mean that we always have to wait for the ready event before actually trying to send a command when connected to sentinels? sorry if its configuration is not properly detailed I will try to do so soo
If condition are always evaluated as true when containing expression syntax inside
Describe the bug An if condition in a step definition that includes expression syntax in condition body is always evaluated as false positive or is ignored. Example workflow: https://github.com/dlavrenuek/test-workflow-if-condition/blob/master/.github/workflows/test.yml [code block] Example run: https://github.com/dlavrenuek/test-workflow-if-condition/runs/2914408827?check_suite_focus=true Example step definition: [code block] In the example workflow the last two steps should not have been run, but they were To Reproduce Steps to reproduce the behavior: 1. create a workflow 2. add an `if` condition with expression syntax `${{ ... }}` as part of the condition (not wrapping the whole condition) 3. trigger the workflow Expected behavior - [ ] The if condition should be correctly evaluated. - [ ] In case that the expression syntax is not supported as part of an `if` condition - an error should be thrown instead of silently ignoring it. Runner Version and Platform GitHub platform What's not working? The steps are executed because of false positive evaluation of he if condition containing expression syntax <img width="541" alt="Bildschirmfoto 2021-06-25 um 15 31 17" src="https://user-images.githubusercontent.com/20122620/123432628-df378580-d5ca-11eb-8f0a-8f2bfa7a488d.png">
Net module based proxy
It look's like @substack implemented a first pass at doing the tcp/net based http proxying ( which calls ry's http parser directly). We should pull this in or use it as a reference for implementing the fast proxying scenarios we were discussing ( not needing the whole request, but just enough to route it ). https://github.com/substack/bouncy/blob/master/index.js
ioredis does not detect dead connections
I am using redis as a cache in a production environment. My redis-server is running on a different VM than my client running ioredis. My problem is that ioredis does not properly handle connections that are "lost" (by which I mean connections that are dead, but have not been properly closed). When this happens, the call-backs of all my .get, .set, etc... calls are only reached after about 11 minutes. For example: [code block] As you can imagine, this is unacceptable behavior for my cache as all the code accessing the cache will get stuck and all requests to my server will timeout. The 'close', 'end' and 'error' events of ioredis are not called, because the tcp connection is not properly closed (no "FIN" tcp packets are sent). This is a common occurrence in a production environment and occurs in the following situations, among others: - the VM running redis-server is shut down - the NAT running between the client and the redis-server (e.g. a firewall or docker's native NATing) eliminates the connection because it has been inactive for some time. This does not happen if you run the client and redis-server on the same machine and just kill (either SIGINT or SIGKILL) the redis-server, as the OS will perform some clean-up and properly close the tcp connections between the redis-server about to be turned off and all connected clients (the ioredis object receives a 'close' event). Coming back to the first situation where the connection is dead because of e.g. a failure of the
Retrive balance on managed account
When I list transaction history on a managed account, I see a list of transaction data. In each of the transaction object I see the value of the "source" attr that start with "py". The transaction object looks like this: `{ "id": "txn_17bBwe2eZvKYlo2Cuwcyi9or", "object": "balance_transaction", "amount": 400, "available_on": 1455235200, "created": 1454687788, "currency": "usd", "description": "Charge for test@example.com", "fee": 42, "fee_details": [ ], "net": 358, "source": "py_17bBwe2eZvKYlo2Crk3VGEG8", "sourced_transfers": { "object": "list", "data": [ [code block] }, "status": "pending", "type": "charge" }` I want to retrieve the detail from the value come from "source", how can I do that? I tried retrieve from Charge,Balance but no luck.
Callbacks Stop Happening
I've been using `mongoose` for a few years, but lately I'm having a lot of trouble where the connection seems to disappear. Connection Settings [code block] The Problem Mongoose just stops calling back. It doesn't emit any errors, memory/cpu is fine on my server & db, and I can connect from any other server. Restarting always fixes it: [code block] Versions On `mongoose 4.4.11`, `mongoose.connection.readyState === UNAUTHORIZED` when callbacks stop. So I upgraded to `mongoose 4.6.0` but within 4 hours the callbacks stopped again, now with `mongoose.connection.readyState === CONNECTED`. Any Ideas Any ideas how I can get this to stop happening? I'm happy to gather any more debugging info, just let me know what info to get.
Question: why is pgvector skipping vector dimenstion check when `typmod` is `-1`
Hi, thanks for this amazing extension! Looking at the source code, I noticed that the dimension check will be skipped if the `typmod` argument of the text input function is `-1`, I am thinking, shouldn't this `typmod` be always given and the check of the input value be always performed? 🤔
Possible memory leak
So there could be a possible memory leak on Express v4.x We have been in constant communications with the Google App Engine team to troubleshoot our servers crashing thought to high memory usage and appears to be a memory leak in Express. Our API is served by Express 4.14.0, the screenshot aboce is an instance only receiving 6 health checks from Google every 5 seconds and within a few days, it crashes because of high memory usage. We thought that was a problem within our base code but we decided to throw a Hello World from Google App Engine which is using only a dependency which is Express 4.15.4 and the same result (graph below). Both instances received health checks from Google every 5 seconds to report their up time and send CPU usage, memory usage, disk usage to Stackdriver The Google Cloud support team was able to reproduce the issue outside of the GCP with the Hello World app and saw the heaps stack size staying consisitent but the overall process memory growing. Since the only dependencies we were using was express we have decided to open an issue. Here is the memory graph that they had. We are currently using Node v6.9.11 in our production.
punycode Deprecation Warning via node-fetch Dependency
Describe the bug When using the openai package in my Node.js project, I encounter the following deprecation warning: After investigating, I found that the openai package relies on node-fetch@2.7.0, which in turn depends on whatwg-url@5.0.0. The whatwg-url package is causing the warning as it depends on the deprecated punycode module. To Reproduce 1. Install openai@latest in a Node.js project. 2. Run any script importing the openai package. 3. Observe the deprecation warning related to punycode. Code snippets _No response_ OS Fedora Node version 22.11.0 Library version 4.73.1
SDK returns empty text blocks, but does not accept them
Using anthropic sdk via langchain_anthropic integration [code block] The SDK can return empty text blocks from models: [code block] However, the SDK does not accept empty text block if re-interacting with the model (e.g., if using an agent).
Catch ECONNREFUSED exceptions
Hi, I want to test what happens to my application when redis goes down. So after starting the server I manually stop the redis daemon that's running on my machine, and I expect to have the errors reported on my error callback/promise. Instead, I get uncaught exceptions all over the place. I've tried using a custom `retryStrategy`, I've tried to set `Redis.Promise.onPossiblyUnhandledRejection` as well... all with no luck. Is there something I'm missing here? How am I supposed to test what happens to my server if for any reason redis stops working? Thanks
[shared_preferences] Android exception: `Unable to establish connection on channel: "dev.flutter.pigeon.shared_preferences_android.SharedPreferencesApi.getAll"`
Long explanation of what is happening https://github.com/flutter/flutter/issues/153075#issuecomment-3313363609 TLDR: Update your Android Gradle Plugin version to 8.9.1 or higher and ensure that all of your android plugins are using at least AGP 8.9.1 or higher. If you have a reproduction that is using AGP 8.9.1 and higher for an app and all of its plugins please add a comment with your reproduction. Steps to reproduce 1. Use old api of shared preferences. 2. Build the app. Expected results It should work with old api of shared preferences. Actual results Getting exception as below Unhandled Exception: PlatformException(channel-error, Unable to establish connection on channel: "dev.flutter.pigeon.shared_preferences_android.SharedPreferencesApi.getAll"., null, null) Code sample <details open><summary>Code sample</summary> [code block] </details> Screenshots or Video <details open> <summary>Screenshots / Video demonstration</summary> [Upload media here] </details> Logs <details open><summary>Logs</summary> [code block] </details> Flutter Doctor output <details open><summary>Doctor output</summary> [code block] </details>
stripe.charge.list missing metadata
When I called `stripe.charges.list(filter, callback);` I get a full list of charges complete with metadata. However, when i call `stripe.charges.list(filter, {stripe_account: accountId}, callback);` to get a subset of charges from the managed account, I am getting the charges without the metadata. How can I get the metadata in the second example?
S3 - RequestTimeout during large files
I'm trying to upload a large file (9 GB) and getting a RequestTimeout error using `aws s3 mv ...` I haven't fully tested it yet, but it seems like if I run the command over and over it will eventually work. Here's the debug log from a failed attempt: https://s3.amazonaws.com/nimbus-public/s3_backup.log I'll post back if I determine that retrying the command several times works or not. aws version: aws-cli/1.1.2 Python/2.7.3 Windows/2008ServerR2
https proxy: socket hang up
I'm getting this error when using https: An error has occurred: {"stack":"Error: socket hang up\n at CleartextStream.<anonymous> (http.js:1272:45)\n at CleartextStream.emit (events.js:61:17)\n at Array.0 (tls.js:617:22)\n at EventEmitter._tickCallback (node.js:126:26)","message":"socket hang up"} When running http it all works just fine. Someone else has recently had a similar problem and was able to solve it (not sure if this would apply here though): http://stackoverflow.com/questions/5680543/nodejs-https-request-for-oauth2-token-google-facebook-produce-error-socket-ha