Repository Issue Activity (beta)

#8081 Daily ironclaw failure taxonomy — 2026-09-07

#8074 Paired user's rejected action in a not-connected shared channel gets the pairing notice copy instead of channel-not-connected copy

#7955 Telegram personal-account linking shows a generic "Something went wrong" when the admin has not configured api_id/api_hash

#7956 Telegram: unpaired sender's /start gets the command inventory instead of the connect/pairing notice

#8066 Prevent command result cards from collapsing when results accumulate

#8065 Align command metadata consistently in the slash-command menu

#8064 Add a dismissal action for command result cards

#8063 Keep the active command visible while navigating the command menu

#7903 Decision spike: persistent per-user sandboxed executor behind the trusted host kernel

#8057 Prompt budget should account for non-transcript prompt material (identity, skills, tool schemas)

#8052 Daily ironclaw failure taxonomy — 2026-09-03

#8033 Remove Redundant `@ts-nocheck` Directives and Prevent New Suppressions

#8036 Type WebUI Test Infrastructure and Remove Remaining Test Suppressions

#8009 MCP egress errors flatten to "response_error", making discovery failures undiagnosable

#8035 Remove `@ts-nocheck` from WebUI Production Components and Hooks

#8041 A tool failure whose kind is wrong sends the model somewhere it cannot recover

#8018 Replace native SettingsField controls with shared Input and SelectMenu

#8034 Add Shared API Types and Remove `@ts-nocheck` from Frontend Boundary Modules

#8032 Eliminate `@ts-nocheck` Debt from the WebUI v2 Frontend

#8017 Adopt shared form and feedback components in Extension Configure

#8020 Use the shared SearchField for Workspace and Logs filters

#8019 Migrate Automations status banners to InlineNotice

#7921 perf(llm): OpenAI-family backends send no prompt_cache_key — measured ~82%→29% cache-hit collapse past ~200 calls

#7971 feat(webui): render model capability tags across Inference selectors

#7970 feat(llm): preserve NEAR AI model modalities through model discovery