Repository Issue Activity (beta)

#11532 Publish GDK crates to crates.io, Maven Central, and PyPI on tag push

#11377 Interactive OAuth sign-in for Azure AI providers without Azure CLI

#11530 OpenRouter: use provider-reported context_length at session time instead of relying solely on the build-time canonical catalog

#11142 Reasoning models produce empty content: max_tokens shared between reasoning_content and content

#11270 CLI: four places disagree about what a slash command is, so Tab hides real commands and bare ones go to the model

#8927 Binaries in git repo make clone slow

#11227 Recipe validation is inconsistent across three paths: manage_schedule reports "Invalid YAML" for valid YAML, and schedule add does not validate at all

#11488 Add option to print a bell when user input is required

#11318 Goose AI Desktop not auto-compacting when hitting context limits with LM Studio

#11526 A goose session cannot recall work done in another agent

#11306 feat(acp): support session-scoped tool compatibility mode

#10682 [BUG] xAI OAuth (SuperGrok) always attempts loopback flow first — silently blocks for the full 5-minute timeout before falling back to device-code, no way to force device-code directly

#11529 Unable to delete chats in Desktop (windows)

#10402 When using a long conversation or a large context window with a local model, Goose attempts to compact the conversation history but fails, rendering the chat unusable. The interface continuously loops and displays the following error message

#11392 Changing extensions via goose still requires a restart/reload of the session

#11253 Extending Goose: Making the GRC More Modular

#11528 When requesting permission to run, the allow/deny box sometimes becomes unclickable

#10955 ACP: disabled extensions are invisible to external harness sessions — no on-demand activation path outside the Goose agent loop

#11069 Anthropic (and other providers') pre-key model list drifts because recommended_models_from_registry() exists but has no callers

#11506 Listing all models available by current provider to fast switch.

#11509 Support cached-input pricing for custom providers

#11524 Telegram Gateway - where?

#11287 perf: session-list query scans all messages with per-row json_extract (>1s on every call)

#11271 Add TrustedRouter as a declarative provider

#11513 Switching the model mid-session pins the chat in Streaming forever (composer dead, backend completes normally)