problem · Revision 2
## Observed symptom cloudflare_api_shield_operation: [Provider] produced an unexpected new value: .with_schemas: was cty.False, but now null. - [x] This is a bug with an existing resource and is not a feature request or enhancement. Feature requests should be submitted with
problem · Revision 3
## Observed symptom Provider panics when pod_failure_policy.rule.on_exit_codes omits container_name (kubernetes_job_v1 / kubernetes_cron_job_v1) Terraform Version, Provider Version and Kubernetes Version `Terraform version: OpenTofu (reproduces independent of CLI; the panic is in
problem · Revision 4
## Observed symptom (bug): `Configuration.alert` panics and recovers due to misconfigured external name function Creation of any `Configuration.alert` panics and recovers with the following provider logs: 2026-10-02T06:27:19Z DEBUG provider-mongodbatlas Cannot observe external re
problem · Revision 2
## Observed symptom [Bug]: `vault_generic_secret` rewrites explicit KV v2 metadata paths to `data/metadata` Terraform Core Version Terraform Vault Provider Version ## Literal error text ``` Error: no secret found at "kv-v2/metadata/apps/example" ``` ## Environment GitHub hashicor
problem · Revision 2
## Observed symptom terraform plan fails: VPC list 404 + backup schedule 404 + nil pointer crash across v2.28.0 and v2.32.0 Terraform plan fails on all recent provider versions (v2.28.0 to v2.32.0) due to three bugs in resourceVultrInstanceRead. This blocks
problem · Revision 3
## Observed symptom `octopusdeploy_tag` create panics (nil pointer) when the tag set update is rejected, hiding the server's error When the server rejects the tag set update during `octopusdeploy_tag` create, the provider crashes instead of reporting the server's error. For
problem · Revision 2
## Observed symptom [BUG] Dashboards Deployment selector built from all `spec.dashboards.labels` The operator copies every entry of `spec.dashboards.labels` into the `spec.selector.matchLabels` of the generated OpenSearch Dashboards Deployment. Since a Deployment's selector is im
problem · Revision 3
## Observed symptom fix(bootstrap): k8s-gate's 'helm dependency build || true' swallows a helm crash as 'no dependencies' <!-- story-spec:prose:start --> Problem The shipped IaC gate template, `development/skills/bootstrap/templates/iac/scripts/k8s-gate.zsh.tmpl`, runs this in it
problem · Revision 2
## Observed symptom [BUG] Helm Secret stringData values are rendered without quotes - [x] I have searched [existing issues](https://github.com/liketrek/TREK/issues) and this bug has not been reported yet - [x] I am running the latest available version of TREK - [x]
problem · Revision 3
## Observed symptom Docker entrypoint fails on Windows checkout due to CRLF line endings Thanks for releasing this project! helped me visualize and track the downloads from an OSS project I'm working on. Bug: On Windows 11 using Git for
problem · Revision 2
## Observed symptom Harness crashes in the MCP reconnection path when a registered MCP server goes down (standalone v0.2.1) ## Environment GitHub issue truefoundry/trueforge #878 ## Trigger / context Public GitHub bug report: MCP reconnection path crash when registered server
problem · Revision 3
## Observed symptom Bundled @ai-sdk/openai-compatible is a pre-3.0.0 copy — still throws "Expected 'function.name' to be a string" (fix exists upstream; interim plugin) Custom `@ai-sdk/openai-compatible` providers still abort turns with `AI_InvalidResponseDataError: Expected 'fun
problem · Revision 2
## Observed symptom Unable to pass serde to PostgresSaver in langgraph.checkpoint.postgres via function from_conn_string Checked other resources - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue.
problem · Revision 3
## Observed symptom Agentic run with dynamic filtering (web_search_20260209 inside code_execution) rejected with "code_execution tool use ... without a corresponding code_execution_tool_result block - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive
problem · Revision 2
## Observed symptom Bug: a Python node's timeout drops LangGraph runnable context Follow-up to [#474](https://github.com/sheikkinen/yamlgraph/issues/474), specifically [the request for a timeout-path reproduction](https://github.com/sheikkinen/yamlgraph/issues/474#issuecomment-58
problem · Revision 2
## Observed symptom bug: Cursor MCP sessions omit capabilities.roots Diagnosed during live testing of the Cursor MCP connection used with this server. Present as a client limitation on `main` at `a3470f7`; this repository only consumes roots. Related server handling: #249.
problem · Revision 3
## Observed symptom RUNTIME Internal error when running cursor-composer exec inside Codex sandbox Repro steps 1. In a Codex session configured with sandbox mode (workspace-write, writes limited to the workspace and /tmp), run: acpx --verbose --format quiet cursor-composer exec 'W
problem · Revision 2
## Observed symptom Cursor inline tool calls replay locally after turnEnded, causing unrequested continuation Cursor executes tools inline, but interaction announcements lack execution ownership. After `turnEnded(stop)`, the agent loop dispatches them again and requests another m
problem · Revision 2
## Observed symptom [Bug]: MCP gateway flattens tool results to text, dropping embedded resources, images and structuredContent - [x] I have searched existing issues and discussions to avoid duplicates - [x] I am using the latest version (or have tested
problem · Revision 2
## Observed symptom Unobserved MCP HTTP transport exception (502) is reported as a daemon crash, but the daemon keeps running Summary An exception from an HTTP MCP server transport (a 502 Bad Gateway response) goes unobserved, and the daemon then