| тред | лейн | касаний | ответил |
ч до ответа | сигналы | заголовок |
|---|
| BerriAI/litellm#38535 | — | 2 | ✅ paoloantinori | 16.8 | | Router policy knob `treat_finish_reason_as_failure`: let a t |
| BerriAI/litellm#34140 | — | 5 | ✅ DanBrima | 22.7 | @ | v3 rate limiter double-counts team per-model limits (model_p |
| BerriAI/litellm#41395 | — | 2 | ✅ DanBrima | 130.1 | @ | [Bug]: context-management summary subrequests bypass a proje |
| BerriAI/litellm#34734 | — | 1 | ✅ aqeeb-warraich | 33.6 | | [Feature]: Add upstream quota and balance probing |
| BerriAI/litellm#34281 | — | 2 | ✅ luckylinux | 0.2 | | [Bug]: Fail Gracefully for Health Checks |
| Jason-Vaughan/TangleClaw#1400 | — | 2 | ✅ Jason-Vaughan | 17.8 | @ | [bug] Primary-checkout guard hook fails silently due to || t |
| JordyZomer/lemmalog#3 | — | 1 | ✅ JordyZomer | 278.3 | | No retraction tool |
| MMoMM-org/the-custom-startup#85 | — | 5 | ✅ MMoMM-org | 10.2 | | agents: evaluate subagent persistent memory (memory: frontma |
| MMoMM-org/the-custom-startup#118 | — | 2 | ✅ MMoMM-org | 0.1 | @ | statusline: use rate_limits from stdin and drop the ccusage |
| NousResearch/hermes-agent#88371 | — | 3 | ✅ liuhao1024 | 1.0 | | [Bug]: `hermes update` crashes with ImportError after " |
| NousResearch/hermes-agent#88378 | — | 1 | ✅ teknium1 | 276.8 | | fix(update): purge stale modules before dashboard cleanup on |
| OpenHands/OpenHands#16308 | — | 1 | ✅ chenhz01 | 297.0 | | [Feature] Usage metrics dashboard for token, ACU, and cost t |
| TIGER-AI-Lab/ClawBench#242 | — | 1 | ✅ Perry2004 | 605.4 | | Judge validation: publish human-agreement (κ) + multi-judge |
| UKGovernmentBEIS/inspect_ai#5413 | — | 1 | ✅ IvanSFlowGit | 25.2 | | Tag empty completions with reason=no_response across the tex |
| UKGovernmentBEIS/inspect_ai#4782 | git-s9 | 1 | ✅ dragonstyle | 26.6 | | fix(solver): prevent closure model retention across calls in |
| UKGovernmentBEIS/inspect_ai#5261 | — | 1 | ✅ dragonstyle | 47.0 | | Fix math() scoring a symbolic answer differently depending o |
| UKGovernmentBEIS/inspect_ai#5266 | — | 1 | ✅ he-yufeng | 15.0 | | fix(scorer): normalize symbol assumptions before math() equi |
| UKGovernmentBEIS/inspect_ai#5090 | — | 1 | ✅ dragonstyle | 107.0 | | fix(solver): support tuples and sequences of solvers in chai |
| UKGovernmentBEIS/inspect_ai#5125 | — | 1 | ✅ epatey | 21.3 | | Fix a sandbox service permanently ceasing to answer requests |
| UKGovernmentBEIS/inspect_ai#4883 | git-s7 | 2 | ✅ nix-aisi | 18.9 | @ | Record Task(approval=...) policies in EvalSpec (#4881) |
| UKGovernmentBEIS/inspect_ai#4928 | git-s9 | 7 | ✅ ClaireXi99 | 3.7 | @ | fix(scorer): map numeric custom values in value_to_float ins |
| UKGovernmentBEIS/inspect_ai#5029 | git-s9 | 1 | ✅ dragonstyle | 48.6 | | fix(scorer): return inf on OverflowError in perplexity metri |
| Wei-Shaw/sub2api#6176 | — | 1 | ✅ shanlilan | 200.7 | | google pro 个人授权提示403 |
| YoanWai/agent-manager#292 | — | 1 | ✅ YoanWai | 363.3 | | TestRefreshNotifiesWaitingTransitionOnce times out on CI |
| accidental-hedge-fund/agent-pipeline#824 | — | 1 | ✅ comamitc | 31.4 | | perf(run-store): scope run-directory scans by issue prefix a |
| agentsmd/agents.md#232 | git-s7 | 2 | ✅ ppseprus | 254.4 | @ | Complementary convention: `agentaccess.txt` — whether agents |
| agno-agi/agno#6892 | — | 1 | ✅ qy6on | 55.4 | | Feature: Verifiable action receipts for agent tool calls |
| agno-agi/agno#10139 | — | 1 | ✅ BrajamohanDas-afk | 1.3 | | [Bug] Curator.prune() and Curator.deduplicate() silently ret |
| agno-agi/agno#10088 | — | 4 | ✅ rodrigocoliveira | 46.1 | @ | fix: exclude the run being continued from its own history in |
| agno-agi/agno#10366 | — | 2 | ✅ FurkhanShaikh | 7.4 | @ | `Agent(retries=N)` re-executes tool calls that already succe |
| agno-agi/agno#10086 | — | 2 | ✅ rodrigocoliveira | 17.3 | @ | [Bug] Background+streaming continue of a paused `external_ex |
| agno-agi/agno#9633 | — | 3 | ✅ chiruu12 | 39.6 | | fix: stringify non-string enum values in Gemini convert_sche |
| anthony-chaudhary/fak#10504 | — | 1 | ✅ anthony-chaudhary | 61.6 | | fix(watchdog): move recovery-critical scheduled tasks off in |
| anthropics/anthropic-sdk-python#1922 | git-s9 | 2 | ✅ adhavan18 | 2.7 | | fix(streaming): wrap mid-stream transport errors as APITimeo |
| anthropics/anthropic-sdk-python#1890 | — | 1 | ✅ okxint | 580.5 | | fix: treat empty-string ANTHROPIC_API_KEY / ANTHROPIC_AUTH_T |
| anthropics/anthropic-sdk-python#1815 | git-s5 | 2 | ✅ sylvesterkaczmarek | 547.9 | | fix(streaming): initialize usage when message_start omits it |
| anthropics/anthropic-sdk-python#1871 | — | 1 | ✅ hipvlady | 300.8 | | `setup_skills` re-downloads and `rmtree`s skills on every se |
| anthropics/anthropic-sdk-python#1927 | — | 2 | ✅ nkato | 5.4 | @ | fix(vertex): serialize the body static-fields-first so promp |
| anthropics/anthropic-sdk-python#1915 | git-s5 | 2 | ✅ iam-kira | 102.4 | @ | Three symlink tests in test_agent_toolset.py are missing the |
| anthropics/anthropic-sdk-python#1820 | git-s5 | 6 | ✅ PiedPiper911 | 42.5 | @ | Fix streaming accumulator crash when message_start omits usa |
| anthropics/anthropic-sdk-python#1906 | — | 5 | ✅ Kayvan-Zahiri | 1.4 | | fix(memory): compare resolved paths in the delete and rename |
| anthropics/anthropic-sdk-typescript#1174 | — | 3 | ✅ AbhiPra24 | 6.0 | @ | fix: preserve defs when schema root is a ref, and fix compac |
| anthropics/anthropic-sdk-typescript#1165 | git-s9 | 6 | ✅ lx3133584 | 0.5 | | fix(BetaMessageStream): preserve null content and existing c |
| anthropics/claude-agent-sdk-python#1276 | — | 2 | ✅ harshal-96 | 8.7 | | fix: parse image content blocks instead of silently dropping |
| anthropics/claude-agent-sdk-python#1271 | — | 2 | ✅ mikevillari | 3.1 | | Expose user-message attribution and queue metadata |
| anthropics/claude-agent-sdk-python#1260 | — | 1 | ✅ danielgines | 11.2 | | Add env parameter to session listing and mutation helpers |
| anthropics/claude-agent-sdk-python#1274 | — | 2 | ✅ feiiiiii5 | 28.1 | | fix(sessions): pick the first JSON field match by position, |
| anthropics/claude-agent-sdk-python#1261 | git-s9 | 2 | ✅ alanhuangyoo | 5.6 | @ | fix(transport): keep reading stderr until the CLI exits on c |
| anthropics/claude-agent-sdk-python#1250 | git-s9 | 3 | ✅ dharaneeshexe-web | 101.4 | | fix: forward strict tool-use flag through tools/list for SDK |
| anthropics/claude-agent-sdk-python#1262 | git-s9 | 1 | ✅ 00200200 | 80.9 | | fix: isolate in-memory session store snapshots |
| anthropics/claude-agent-sdk-python#1246 | — | 8 | ✅ 1fanwang | 1.3 | | Fix raw asyncio cancellation leaking CLI subprocesses |
| anthropics/claude-agent-sdk-python#1234 | — | 1 | ✅ yohimik | 397.3 | | Bug: claude-agent-sdk 0.2.144 is a partial publish |
| anthropics/claude-agent-sdk-python#1237 | git-s9 | 3 | ✅ rioyu123 | 3.1 | 👍1 | fix(sessions): surface subagent import I/O failures |
| anthropics/claude-code#82056 | git-s9 | 11 | ✅ DanceNitra | 0.9 | @ | A session cannot determine whether its auto-memory index loa |
| anthropics/claude-code#95833 | — | 2 | ✅ Alpha-Oi | 37.5 | @👍1 | [BUG] PreToolUse hooks (Bash/PowerShell matchers) never fire |
| anthropics/claude-code#93015 | — | 1 | ✅ jessemi11er | 30.0 | | Scheduled tasks stamp lastRunAt but never launch a session ( |
| anthropics/claude-code#83323 | — | 1 | ✅ GeografiaenDatos | 873.2 | @ | Scheduled tasks silently stop firing after first run — recre |
| anthropics/claude-code#91188 | git-s7 | 11 | ✅ pm25coder | 2.2 | @ | Feature request: make the auto-memory MEMORY.md compaction r |
| anthropics/claude-code#90542 | — | 1 | ✅ stonianua | 0.6 | @ | [BUG] A complete CLAUDE.md rule contract governed nothing: 9 |
| anthropics/claude-code#89040 | — | 2 | ✅ ehawkin | 51.1 | | /compact silently fails to apply on very large conversations |
| anthropics/claude-code#70555 | — | 1 | ✅ kyle641320 | 356.5 | @ | Working-state continuity: survive compaction and /clear (the |
| anthropics/claude-code#91870 | — | 1 | ✅ AlmogBaku | 3.0 | | Mods - make Claude 10x more extensible |
| anthropics/claude-code#88178 | — | 3 | ✅ rhubain | 22.9 | | Claude Desktop: 5h46 of cumulative waiting in one day — 23 s |
| anthropics/claude-code#2544 | — | 1 | ✅ durban01 | 336.5 | | [BUG] CLAUDE.md Mandatory Rules Consistently Ignored Across |
| anthropics/claude-code#91884 | — | 1 | ✅ JohnnyRayWork | 251.7 | | Desktop scheduled tasks: model selection broken end-to-end ( |
| anthropics/claude-code#82134 | — | 2 | ✅ mrsoone | 557.2 | @ | Windows MSIX app: auto-update during app hang corrupts packa |
| anthropics/claude-code#91371 | — | 2 | ✅ lululin221010 | 15.6 | | [BUG] Local scheduled tasks silently hang mid-run and block |
| anthropics/claude-code#69364 | — | 1 | ✅ gmantova | 345.2 | | --continue / --resume doesn't check the existing live-s |
| anthropics/claude-code#93231 | — | 1 | ✅ JohnRSandoval | 305.7 | | Session exit on VS Code window close never releases its git |
| anthropics/claude-code#80773 | — | 1 | ✅ flober81 | 121.2 | | [BUG] claude://resume duplicates a session already open as a |
| anthropics/claude-code#76357 | — | 2 | ✅ AccountZero | 54.6 | | [BUG] Windows (MSIX): update fails with 'Another progra |
| anthropics/claude-code#94013 | — | 2 | ✅ danpingsmile1197-design | 17.4 | | Background subagents have no token, turn or time cap: three |
| anthropics/claude-code#81081 | — | 5 | ✅ DanceNitra | 7.9 | @ | [BUG] Session-start skill listing silently truncates most sk |
| anthropics/claude-code#38335 | — | 1 | ✅ fgrosswig | 1.9 | @ | [BUG] Claude Max plan session limits exhausted abnormally fa |
| anthropics/claude-code#92583 | — | 1 | ✅ wshallwshall | 325.7 | | Windows: Bash tool commands auto-backgrounded on timeout are |
| anthropics/claude-code#95582 | — | 1 | ✅ Getgodmode | 6.5 | | Skill catalog descriptions intermittently missing from syste |
| anthropics/claude-code#85422 | — | 1 | ✅ mycroft | 0.3 | @ | [FEATURE] Token-burn circuit breaker: runtime-enforced spend |
| anthropics/claude-code#90349 | — | 1 | ✅ ghbaud | 0.3 | | Feature request: let a session start a new named session in |
| anthropics/claude-code#92505 | — | 1 | ✅ boydenus | 7.5 | | Model reports work as completed that it did not do, at a vol |
| anthropics/claude-code#86915 | — | 2 | ✅ bryandent | 383.8 | @ | Scheduled task session hangs forever on unattended permissio |
| anthropics/claude-code#84793 | — | 1 | ✅ clementrahula | 215.0 | | remoteControlAtStartup not honored when the desktop app resu |
| anthropics/claude-code#90215 | — | 1 | ✅ pez-stack | 195.8 | | [BUG] Scheduled task cron dispatch silently fails to fire — |
| anthropics/claude-code#47180 | — | 3 | ✅ mattcolquhoun | 59.7 | | [BUG] Cowork scheduled tasks ignore "Always allow" |
| anthropics/claude-code#90077 | — | 1 | ✅ Suyann | 155.9 | | [BUG] Hooks with `shell: "powershell"` spawn `pwsh |
| anthropics/claude-code#77817 | — | 1 | ✅ rodolfopaiz | 237.2 | | [BUG] v2.1.206 silently broke unattended scheduled tasks: ru |
| anthropics/claude-code#92825 | — | 1 | ✅ ANGELBERRIOS23 | 10.6 | @ | [BUG] Claude Code Desktop: session transcripts silently beco |
| anthropics/claude-code#93046 | — | 2 | ✅ mhammadjaber00 | 0.2 | | Usage-limit warning names the parent model, not the subagent |
| anthropics/claude-code#77657 | — | 1 | ✅ kazuki-uike-coconala | 507.7 | | [BUG] Scheduled Tasks: lastRunAt/nextRunAt inconsistent with |
| anthropics/claude-code#93260 | — | 1 | ✅ TheAviv | 102.6 | | Projects: concurrent sessions silently clobber each other&#x |
| anthropics/claude-code#92264 | — | 1 | ✅ korczis | 134.3 | | Idle background session stops advancing while async subagent |
| anthropics/claude-code#66402 | — | 4 | ✅ abhideepm | 21.1 | @ | [BUG] /model and /effort mutate global settings.json — break |
| anthropics/claude-code#92998 | — | 1 | ✅ hemangnagar | 13.8 | | Auto-memory: MEMORY.md overflow silently discards the NEWEST |
| anthropics/claude-code#85873 | — | 2 | ✅ luckylucab0 | 0.7 | @ | [BUG] Scheduled tasks created inside an SSH remote session n |
| anthropics/claude-code#92429 | — | 1 | ✅ aurumflux20 | 70.3 | | Scheduled task marks itself as completed (lastRunAt/nextRunA |
| anthropics/claude-code#30519 | — | 3 | ✅ m13v | 245.9 | | Permissions matching is fundamentally broken — 30+ open issu |
| anthropics/claude-code#81833 | — | 6 | ✅ daichiyasunami-vottia | 0.1 | @ | Auto-memory is inconsistently loaded in git-worktree session |
| anthropics/claude-code#78569 | git-s5 | 5 | ✅ daichiyasunami-vottia | 0.1 | @ | [BUG] Auto-memory instructions direct an immediate MEMORY.md |
| anthropics/claude-code#92563 | — | 1 | ✅ sergey-gusev94 | 0.3 | | Completed background-task notification is appended to the se |
| anthropics/claude-code#91905 | git-s5 | 3 | ✅ stonianua | 5.5 | | Fable 5.1 repeatedly ignores a standing "read the recor |
| anthropics/claude-code#81992 | — | 6 | ✅ alfredoramses | 0.2 | @ | Claude Desktop repeatedly enters "Modified, NeedsRemedi |
| anthropics/claude-code#82084 | — | 2 | ✅ BasedGPT | 28.0 | | Session transcripts are silently and permanently deleted aft |
| anthropics/claude-code#77001 | — | 1 | ✅ tonydehnke | 0.7 | | Disproportionate subscription usage burn in long sessions: c |
| anthropics/claude-code#85617 | — | 1 | ✅ Octonove | 573.8 | | Model asserted third-party delivery worked from sender-side |
| anthropics/claude-code#43698 | — | 1 | ✅ nta76 | 486.5 | | [FEATURE] Multi-device Cowork: Allow concurrent Cowork agent |
| anthropics/claude-code#91642 | — | 3 | ✅ runvouch | 1.6 | | Scheduled-task CLI process does not exit after unattended ru |
| anthropics/claude-code#83932 | — | 3 | ✅ Kildere | 33.9 | @ | [BUG] Windows auto-update deploys into running claude.exe + |
| anthropics/claude-code#85199 | — | 1 | ✅ SplendidHowl | 9.7 | | [BUG]Claude Desktop repeatedly crashes and requires “Advance |
| anthropics/claude-code#88982 | — | 1 | ✅ marius-bughiu | 130.4 | @ | Scheduled task sessions (Desktop local agent mode) never exi |
| anthropics/claude-code#84333 | — | 1 | ✅ codebytere-ant | 21.9 | | [BUG] Claude Desktop (Windows MSIX) silently becomes Modifie |
| anthropics/claude-code#89639 | — | 1 | ✅ bakemocho | 92.8 | | macOS: scheduled-task sessions wedge mid tool-call (~30s in, |
| anthropics/claude-code#54461 | — | 1 | ✅ ben-aggus-gateway | 167.4 | | Desktop app: cannot change primary working directory or open |
| anthropics/claude-code#89632 | — | 3 | ✅ runvouch | 1.3 | @ | Local scheduled tasks run under interactive ask-every-tool p |
| anthropics/claude-code#90659 | — | 4 | ✅ trakshan-mishra | 2.6 | @ | Backgrounded Bash commands can report a false "exit cod |
| anthropics/claude-code#79217 | — | 4 | ✅ miguelsambricio-cyber | 32.7 | @ | [FEATURE] Make the auto-memory MEMORY.md index size limit (2 |
| anthropics/claude-code#84851 | — | 2 | ✅ mickaela-stack | 21.7 | | [BUG] Windows MSIX auto-update corrupts package (Modified, N |
| anthropics/claude-code#89283 | — | 3 | ✅ Trocergian | 0.4 | | [BUG] Desktop app creates duplicate project directories for |
| anthropics/claude-code#82546 | — | 1 | ✅ tamilov | 99.2 | | [BUG] /goal set at a compact_boundary never starts its turn; |
| anthropics/claude-code#88138 | — | 1 | ✅ codebytere-ant | 21.9 | | [BUG] Claude Desktop (Windows MSIX) transitions from Ok to M |
| anthropics/claude-code#81180 | — | 1 | ✅ codebytere-ant | 21.8 | | [BUG] Claude Desktop (Windows MSIX) reverts to Modified, Nee |
| anthropics/claude-cookbooks#803 | git-s5 | 13 | ✅ FutureEnterprises | 7.0 | | fix(managed-agents): bind approval to exact merge |
| aws/bedrock-agentcore-starter-toolkit#498 | — | 1 | ✅ pshfrank7 | 356.6 | | No API to list or force-terminate active runtime sessions — |
| bartoszkobylinski/lovspor#246 | — | 1 | ✅ bartoszkobylinski | 2.0 | | /security-check silently skips two source files: uvx bandit |
| basicmachines-co/basic-memory#993 | git-s5 | 4 | ✅ ajdelaguila | 105.0 | | Use frontmatter to enforce validation and set conditional (r |
| basicmachines-co/basic-memory#1430 | — | 1 | ✅ phernandez | 76.1 | | Windows CI: test_chatgpt_search_pagination_default hangs in |
| basicmachines-co/basic-memory#1275 | — | 2 | ✅ phernandez | 246.9 | | file_path lookups are byte-wise, so NFC/NFD filename variant |
| browser-use/browser-use#5767 | — | 1 | ✅ soroush5 | 36.2 | | AgentHistoryList.save_to_file/load_from_file silently drops |
| browser-use/browser-use#5633 | — | 1 | ✅ Mohil-Ahuja | 1.1 | | Bug: AnthropicMessageSerializer drops earlier system messag |
| browser-use/browser-use#5663 | — | 1 | ✅ Mohil-Ahuja | 1.0 | | fix(llm): keep every system message in the Anthropic request |
| browser-use/browser-use#5137 | — | 1 | ✅ Dante-dan | 36.1 | | Proposal: add structured action evidence for no-op and chang |
| browser-use/browser-use#5768 | — | 1 | ✅ soroush5 | 7.9 | | fix(agent): preserve token usage in AgentHistoryList save/lo |
| browser-use/browser-use#5529 | — | 1 | ✅ noma-py | 234.8 | | Bug: failed automatic new-tab switch is reported as successf |
| browserbase/stagehand#2486 | — | 1 | ✅ alectimison-maker | 131.7 | | v4: locator.click() returns without committing a selectable |
| bytedance/deer-flow#5391 | — | 1 | ✅ chenzhiya77 | 27.0 | | RFC: 内置本地知识库（Harness RAG） |
| caura-ai/caura#1467 | git-s7 | 1 | ✅ deepspace28 | 6.0 | | fix(client-typescript): recall() silently drops caller extra |
| cohere-ai/cohere-python#807 | — | 1 | ✅ ProgrammerPlus1998 | 105.4 | | [HTTPXodus] migrate httpx to httpx2 (hard switch; closes #80 |
| comet-ml/opik#8134 | — | 1 | ✅ petrotiurin | 233.0 | | Evaluation pipeline: failed metrics silently dropped from av |
| crewAIInc/crewAI#6043 | — | 1 | ✅ impartshadow | 14.6 | @ | [FEATURE] Memory write guards for multi-agent crews — preven |
| crewAIInc/crewAI#5802 | — | 1 | ✅ giskard09 | 1.1 | | Tool re-execution on task retry has no idempotency guard — d |
| deepset-ai/haystack-core-integrations#3544 | — | 1 | ✅ julian-risch | 578.9 | | Remove the meta-llama integration |
| desimpson/syncer#158 | — | 2 | ✅ desimpson | 139.4 | @👍1 | Bug: Firefox profile auto-detect fails on Windows because pr |
| eirkkr/fieldkit#77 | git-s7 | 1 | ✅ eirkkr | 68.9 | @ | Add writing rules for prose, and a hook to trigger them |
| eugeniughelbur/obsidian-second-brain#171 | git-s5 | 5 | ✅ eugeniughelbur | 0.2 | @👍1 | Own a platform build: all seven are open |
| evalstate/fast-agent#906 | git-s5 | 4 | ✅ AmirF194 | 3.9 | | fix(mcp): track ping error responses as connection failures, |
| fer-osorio/claude-sandbox#65 | — | 1 | ✅ fer-osorio | 1.6 | | docs(global-layer): design curated auto-memory seeding |
| garrytan/gbrain#4798 | — | 1 | ✅ gtdevelopment | 0.8 | | UTF-8 BOM silently breaks heading title extraction (title fa |
| gastownhall/beads#5877 | git-s7 | 9 | ✅ RemanenetSpy | 3.4 | @ | Proposal: Memory Beads |
| get-bb/bb#2433 | — | 1 | ✅ ymichael | 326.7 | | No per-host capacity model: bb over-schedules one machine un |
| get-bb/bb#1552 | — | 1 | ✅ SawyerHood | 117.3 | | Pool multiple Claude accounts: limit-aware load balancing ac |
| getzep/graphiti#1645 | — | 3 | ✅ brentkearney | 86.2 | | MCP server: search behavior is hardcoded — proposal: configu |
| google-gemini/gemini-cli#29101 | — | 1 | ✅ Anonymous12-png | 1.0 | @👍1 | Authentication failure blocking Enterprise Workspace account |
| google/adk-go#1333 | — | 1 | ✅ coldcoder126 | 15.4 | | [Feature]: model/openaimodel: support image/file input (gena |
| google/adk-python#6957 | git-s9 | 7 | ✅ prasanna8585 | 6.2 | @ | fix(auth): stop OAuth2 client_secret and tokens from leaking |
| google/adk-python#7044 | git-s5 | 5 | ✅ claxman | 4.5 | | fix(plugins): wrap class descriptors from __dict__ |
| google/adk-python#6942 | — | 4 | ✅ businessarshgoyal | 19.6 | | fix(sessions): normalize session id on read and delete too |
| google/adk-python#6933 | — | 2 | ✅ gioboa | 27.8 | @ | fix(mcp): recover pooled session after server-side terminati |
| google/adk-python#7030 | git-s5 | 1 | ✅ sanketpatil06 | 35.1 | | Artifact services (in_memory, file, GCS) key storage on raw |
| google/adk-python#6710 | git-s9 | 1 | ✅ gaurav-gandhi-2411 | 375.4 | | fix(evaluation): record NOT_EVALUATED instead of dropping in |
| google/adk-python#6891 | git-s9 | 1 | ✅ DeanChensj | 26.8 | @ | fix(sessions): trim session id before the in-memory duplicat |
| google/adk-python#6729 | git-s8 | 3 | ✅ chelsealong | 3.6 | | fix(sessions): make SqliteSessionService state merges use di |
| huggingface/smolagents#2665 | — | 2 | ✅ fszcd | 23.6 | | Fix color inversion for tensor-backed `AgentImage` |
| huggingface/smolagents#2656 | — | 1 | ✅ fszcd | 134.0 | | [BUG] AgentImage inverts pixel values when built from a tens |
| jamditis/audiobud#107 | — | 1 | ✅ jamditis | 44.9 | | [BUG] Parakeet repeats a letter mid-acronym: "NJ PBS&qu |
| jerryfane/herdr#66 | — | 1 | ✅ jerryfane | 252.0 | @ | design: herdr-per-machine federation over Tailscale (cross-m |
| joshuaswarren/remnic#1962 | — | 5 | ✅ RemanenetSpy | 776.7 | @ | hypothesis(H5): origin-bound authority neutralizes memory in |
| justinmclean/jobhunter#5 | — | 4 | ✅ GregoryBolshakov | 1.3 | | Wanted: new source adaptors (see ADAPTORS.md) |
| kirodotdev/KiroCrew#11562 | — | 1 | ✅ bolichen97 | 19.4 | | New sessions start on the last picker-selected model, not th |
| kunchenguid/firstmate#1818 | — | 1 | ✅ kunchenguid | 3.2 | | Crewmates are cut off from the project's auto-memory: t |
| langchain-ai/langgraph#7417 | — | 1 | ✅ Pushkard17 | 3.1 | | Long tool calls (~180s+) silently re-executed from checkpoin |
| langchain-ai/langgraph#9006 | — | 1 | ✅ AgentEtna-Founder | 2.4 | | Checkpoint resume after tool timeout: resume vs full restart |
| langchain-ai/langgraph#8764 | — | 3 | ✅ mstevens843 | 0.2 | @👍2 | Crash before first durable checkpoint can silently drop an a |
| langchain-ai/langgraph#8039 | — | 2 | ✅ safal207 | 3.5 | @ | durability="sync": put_writes/put persistence orde |
| langfuse/langfuse#15372 | — | 1 | ✅ yagosamu | 197.1 | | bug(sdk-python): mask is silently skipped for third-party OT |
| langfuse/langfuse#16494 | — | 1 | ✅ aniruddhaadak80 | 12.7 | | bug(automations): Execution logs get stuck on PENDING when j |
| macanderson/stella#5287 | — | 1 | ✅ macanderson | 60.4 | | Parallel agent sessions can share one checkout, and a branch |
| malamoney/jobfinder#2 | — | 1 | ✅ malamoney | 0.2 | | Build Jobfinder: nightly Board sweep, Criteria matching, and |
| max-friedman/agentic-coding-loop#23 | — | 1 | ✅ max-friedman | 43.8 | @ | [proposal] Mutation-check the CARRIER, not just the ends — f |
| mem0ai/mem0#5352 | — | 1 | ✅ Ardem2025 | 9.9 | @ | Workaround: Solving time-blindness, semantic conflicts, and |
| mem0ai/mem0#7211 | — | 1 | ✅ nothariharan | 79.6 | | bug(vector-stores): Azure AI Search indexing errors silently |
| mem0ai/mem0#6859 | — | 2 | ✅ iroiro147 | 29.7 | | fix(vector_stores/azure_ai_search): raise on failed indexing |
| mem0ai/mem0#6935 | — | 1 | ✅ kartik-mem0 | 157.8 | | fix(milvus): convert COSINE distance to similarity in _parse |
| mem0ai/mem0#6214 | — | 1 | ✅ infinitus01 | 14.3 | | Feature: portable session export/import on the core Memory S |
| mem0ai/mem0#5867 | — | 1 | ✅ RemanenetSpy | 6.3 | | ADD-only memory extraction can create conflicting memories |
| microsoft/agent-framework#7383 | — | 2 | ✅ babyblueviper1 | 0.1 | | Python: Bind tool-approval responses to surfaced approval re |
| mistralai/client-python#532 | — | 1 | ✅ Nelson-PROIA | 1053.4 | | [BUG CLIENT]: Connection leak: streaming requests that retur |
| mistralai/mistral-vibe#1065 | git-s2 | 4 | ✅ Ritabanm | 3.0 | @ | fix(app_server): increase connector tool limit to 1000 (#105 |
| mixelpixx/Konnect#481 | git-s7 | 3 | ✅ mixelpixx | 29.8 | | fix(drc): name what owns each DRC report item |
| mixelpixx/Konnect#103 | — | 2 | ✅ mixelpixx | 512.0 | | Plugin leaks orphan konnect server processes: PID file only |
| mixelpixx/Konnect#498 | git-s7 | 1 | ✅ neusse | 4.2 | | IPC socket auto-detection wedges KiCad's API server on |
| mixelpixx/Konnect#411 | — | 1 | ✅ neusse | 18.8 | 👍1 | score_placement's decoupling check flags a correctly-pl |
| modelcontextprotocol/go-sdk#1132 | git-s9 | 1 | ✅ yarolegovich | 441.5 | | mcp: prevent MemoryEventStore.After panic on index past the |
| modelcontextprotocol/go-sdk#1095 | — | 1 | ✅ yarolegovich | 570.2 | | No way to restrict the protocol versions a server advertises |
| modelcontextprotocol/typescript-sdk#2706 | git-s9 | 2 | ✅ Gursimrxn | 1.8 | | fix(stdio): share a single drain listener under backpressure |
| monk-io/monk-plugin#367 | — | 2 | ✅ NyxSpecter4 | 10.4 | @ | [Bug bounty] Register-AntigravityMcp writes UTF-8 BOM into m |
| obra/superpowers#2046 | — | 1 | ✅ kylekellogg | 177.8 | | writing-plans can mistake structural RED for behavior-test e |
| open-telemetry/semantic-conventions-genai#487 | — | 3 | ✅ roy-tong | 41.4 | | Usage buckets: how should consumers know which detail attrib |
| open-telemetry/semantic-conventions-genai#403 | — | 3 | ✅ alebgl77 | 15.1 | @ | Record token usage on fetch/poll spans for background execut |
| openai/codex#31592 | — | 1 | ✅ anotherempty | 116.6 | | Skill discovery ignores symlinked SKILL.md files |
| openai/codex#35347 | — | 1 | ✅ thisisntjon | 508.2 | | [Windows] Codex desktop app fails to launch, AppX package st |
| openai/openai-agents-js#1752 | git-s9 | 1 | ✅ pacocartones | 3.5 | | fix(extensions): decode PTY output incrementally |
| openai/openai-agents-js#1775 | — | 3 | ✅ MohammedAlkindi | 0.0 | | code-change-verification cannot run on Windows under Node 20 |
| openai/openai-agents-python#4890 | git-s9 | 1 | ✅ wolfgang-aura | 83.9 | | fix: keep the file when apply_patch does a case-only rename |
| openai/openai-agents-python#4827 | — | 1 | ✅ dixso | 28.5 | | Deferred interrupted-turn session items are never persisted |
| openai/openai-agents-python#5025 | — | 2 | ✅ seratch | 61.8 | | Support skills for Agent and not just for SandboxAgent |
| openai/openai-agents-python#4889 | git-s5 | 2 | ✅ oyiakoumis | 232.6 | | [Sandbox] apply_patch update_file with a case-only move_to d |
| openai/openai-agents-python#5015 | — | 1 | ✅ seratch | 37.6 | | fix(core): keep tool output custom_data structured in RunSta |
| openai/openai-agents-python#5024 | — | 1 | ✅ seratch | 51.1 | | needs_approval misconfiguration raises in the core runner an |
| openai/openai-agents-python#4777 | — | 4 | ✅ Nikhils-G | 0.2 | | VoicePipeline rejects string dtype spellings in TTS settings |
| openai/openai-agents-python#4778 | — | 3 | ✅ Nikhils-G | 0.7 | | fix(voice): accept every NumPy spelling of a supported TTS d |
| openai/openai-agents-python#4797 | — | 3 | ✅ harshitethic | 29.1 | | fix(voice): accept string TTS dtypes |
| patraratorn/agentrouter-claude-proxy#1 | — | 1 | ✅ patraratorn | 1.9 | | install-autostart.ps1 fails: invalid LogonType "Interac |
| pingdotgg/t3code#9439 | — | 3 | ✅ thatmike1 | 1.5 | | fix(usage): keep the largest output count in a Claude dedupe |
| planetf1/otelite#113 | — | 2 | ✅ planetf1 | 288.6 | | feat: context-composition view — decompose cached prefix int |
| pydantic/pydantic-ai#7975 | — | 4 | ✅ roy-tong | 41.4 | | OTel usage attributes: should subset detail keys (reasoning/ |
| raphaelfnds/rsct-framework#57 | — | 1 | ✅ raphaelfnds | 32.7 | @ | feat(mcp): plans carry no value or cost signal — the only or |
| run-llama/llama_index#23122 | — | 1 | ✅ Kaushalt2004 | 0.5 | @ | Integration proposal: memory store with verified transfer — |
| ruvnet/ruflo#2968 | — | 1 | ✅ O96a | 500.7 | | 3.37.0 regression: memory store reports success but persists |
| screenpipe/screenpipe#6661 | — | 1 | ✅ EzraEllette | 308.0 | | [question] Free plan storage block - is this intended? |
| stablyai/orca#21628 | — | 1 | ✅ nwparker | 21.6 | | fix: report missing Antigravity hook scripts |
| stripe/ai#402 | — | 2 | ✅ iwasinnam2 | 2.2 | @ | Agent-level retry creates duplicate charges — no idempotency |
| supermemoryai/supermemory#1348 | — | 2 | ✅ nagasatish007 | 76.7 | | [Feature Request] Memory governance layer — PII redaction, c |
| taylorwilsdon/google_workspace_mcp#946 | — | 1 | ✅ MMannanAkmal | 13.2 | | streamable-http pre-flight port-bind check causes a silent c |
| techjarves/Mobile-Harness#10 | git-s4 | 1 | ✅ techjarves | 108.1 | | BUG - Selecting Claude Subscription in the AI connections do |
| tenequm/pond#166 | — | 2 | ✅ tenequm | 276.3 | | tenants: keep work and personal sessions apart, scope what e |
| thedotmack/claude-mem#3815 | — | 1 | ✅ thedotmack | 245.5 | | Worker daemon has no USER/LOGNAME, so the SDK child falls ba |
| thedotmack/claude-mem#3814 | — | 1 | ✅ thedotmack | 245.5 | | Observer output: constrained decoding via JSON schema remove |
| truera/trulens#2730 | — | 1 | ✅ joshreini1 | 239.3 | @ | fix(database): SQL-level aggregation in pre-OTel leaderboard |
| vansh7nvc/Abstractify#12 | — | 5 | ✅ vansh7nvc | 8.3 | @ | Issue #12 🎙️ Abstract-to-Podcast Audio Summary (Two-Host Str |
| woojubb/robota#2577 | — | 1 | ✅ woojubb | 20.9 | | [enhancement] Add a local cross-session usage dashboard for |
| wowok-ai/skills#3 | — | 3 | ✅ NatureBlueee | 0.4 | @👍1 | Global install left one client with no skills and no MCP reg |
| xCirno1/applyer#29 | — | 3 | ✅ GregoryBolshakov | 1.3 | @ | Search company ATS boards (Greenhouse/Lever/Ashby) as a job |
| xai-org/xai-sdk-python#207 | — | 1 | ✅ GISWLH | 9.6 | @ | fix: use monotonic clock in PollTimer to survive wall-clock |
| xjiang16/job-market-tracker#28 | — | 1 | ✅ xjiang16 | 182.7 | | Ingest company ATS board postings (Greenhouse / Lever / Ashb |
| zilliztech/memsearch#676 | git-s7 | 3 | ✅ ilipkanou | 25.6 | | Claude Code SessionStart hook still hits `timeout: 10` after |
| zilliztech/memsearch#658 | git-s5 | 1 | ✅ sdkkds | 271.6 | @ | Windows/Git Bash: stop_watch orphan sweep is a no-op (no pgr |
| zilliztech/memsearch#600 | — | 1 | ✅ sdkkds | 58.4 | | Windows: kill_orphaned_index can't reap native index pr |
| 1jehuang/jcode#960 | — | 1 | — | — | | Memory retrieval gaps: near-literal search misses, stale mem |
| 51hcie/ai-project-continuity#20 | — | 1 | — | — | | Conflict handling when multiple AI agents concurrently updat |
| Altinn/kihub#117 | — | 1 | — | — | | Specify kill-switch protocol: scope, testing, and sub-agent/ |
| AndreasSteinerPF/team-memory#7 | — | 1 | — | — | | Design an expiry and revalidation workflow for stale memorie |
| BerriAI/litellm#42210 | — | 1 | — | — | | fix(anthropic): scope OAuth token forwarding to official Ant |
| BerriAI/litellm#39370 | — | 1 | — | — | | [Bug]: Reset-budget job never self-heals a budget_duration=n |
| BerriAI/litellm#27470 | — | 1 | — | — | | Cooldown TTL doesn't distinguish 429-rate-limit from 42 |
| BerriAI/litellm#33371 | — | 1 | — | — | | [RFC] Expose structured provider-error/route-health contract |
| BerriAI/litellm#39979 | — | 1 | — | — | | [Bug]: Request Logs date-range filter interprets the picker& |
| CIRISAI/CIRISAgent#907 | — | 1 | — | — | | Memory poisoning: provenance / trust-scoring on retrieved co |
| Eppie-io/Eppie-CLI#622 | — | 1 | — | — | | UTF-8 BOM becomes part of piped vault password |
| GAM-team/got-your-back#333 | — | 1 | — | — | | QuotaLimitExceeded triggered, does not retry, and falsely re |
| Gerico1007/deepdiver#25 | — | 1 | — | — | | Fix NotebookLM audio overview artifact completion and downlo |
| Hal0ai/hal0#1537 | — | 1 | — | — | | install: on Ubuntu 26.04 (python3.14) the Hindsight engine i |
| JeltzProstetnic/agent-fleet#12 | — | 1 | — | — | | SessionStart additionalContext payload silently fails to rea |
| Laticent/lattice#1069 | — | 1 | — | — | | lint: proseWordCount counts fenced-code content as prose (wa |
| Lyellr88/marm-memory#170 | — | 3 | — | — | 👍1 | testing: expand Docker coverage, only 8 of 43 Docker tests r |
| ManikantaR/playlistminer#11 | — | 1 | — | — | | Learning agent: watch-history import via Google Takeout (Pha |
| MemPalace/mempalace#224 | — | 1 | — | — | | Stale drawer retrieval can inject contradictory memory into |
| MichaelYcJo/SpecSeal#129 | — | 1 | — | — | | agents/smith.md's mutation-testing instruction clears t |
| Microck/satelle#199 | — | 1 | — | — | | windows session 0 background service isolation blocks deskto |
| Morrison-Lab/ai-config#2004 | — | 1 | — | — | | Every PreToolUse and Stop hook is inert in remote/web sessio |
| NathanKrupa/OverSteward#287 | — | 1 | — | — | | fix(dream): MEMORY.md is 34KB against a 25KB cap and silentl |
| NishikawaButterfly/quant-risk-engine#51 | — | 1 | — | — | | Catch SystemExit in the CLI test harness |
| NousResearch/hermes-agent#27013 | — | 1 | — | — | | Agents lose project context across session restarts — halluc |
| Oscar-Build/Automating_podcast#57 | — | 1 | — | — | | [M2-2.1] Validate NotebookLM automation route (headless audi |
| PrefectHQ/fastmcp#4899 | — | 1 | — | — | | Client .data returns date, time, timedelta and UUID as strin |
| PrismorSec/prismor#272 | — | 1 | — | — | | Runaway agent loop detection and circuit breakers |
| Skyvern-AI/skyvern#3897 | — | 1 | — | — | | Issue: Cookies Not Persisting Between Workflows in Persisten |
| TencentCloud/TencentDB-Agent-Memory#1025 | — | 1 | — | — | | Proposal: proactive L1 memory maintenance — LLM-driven recal |
| UKGovernmentBEIS/inspect_ai#4881 | git-s7 | 1 | — | — | | Tool Approval is not properly recorded in EvalSpec |
| VibePod/vibepod-cli#116 | git-s4 | 1 | — | — | | IKWID for claude: settings-based bypass instead of --dangero |
| Wan-ZL/zelin-ai-assistant#28 | — | 1 | — | — | | Disk retention: surface screenpipe data usage and retention |
| agent0ai/dox#1 | — | 1 | — | — | | Consider an optional machine-checkable convention for AGENTS |
| agentsmd/agents.md#228 | — | 1 | — | — | | AGENTS.md is being used as a policy file, but two of its own |
| agentsmd/agents.md#185 | — | 1 | — | — | | Need a standardized way to show different AGENTS.md files an |
| agno-agi/agno#7879 | — | 1 | — | — | | Bug: session I/O wrappers in _storage.py swallow DB exceptio |
| agno-agi/agno#9928 | — | 1 | — | — | | [Bug] HITL pause from a custom function-executor Step is dro |
| anthropics/anthropic-sdk-python#1921 | git-s9 | 1 | — | — | | fix(streaming): raise the exception class matching an in-str |
| anthropics/anthropic-sdk-python#1914 | — | 1 | — | — | | fix(memory): require a separator after the /memories prefix |
| anthropics/anthropic-sdk-python#1830 | git-s9 | 2 | — | — | | fix(streaming): accumulate compaction delta content |
| anthropics/anthropic-sdk-typescript#986 | — | 1 | — | — | | partialParse in _vendor/partial-json-parser crashes on inval |
| anthropics/claude-agent-sdk-python#1109 | — | 1 | — | — | | fix: surface prompt stream errors instead of hanging query() |
| anthropics/claude-agent-sdk-python#1108 | — | 1 | — | — | | Exception in the prompt AsyncIterable is swallowed at debug |
| anthropics/claude-agent-sdk-python#1162 | — | 1 | — | — | | CLI version check timeout can hang indefinitely while waitin |
| anthropics/claude-agent-sdk-python#1165 | — | 1 | — | — | | max_buffer_size and import byte limits count characters inst |
| anthropics/claude-code#88579 | — | 1 | — | — | | Persistent memory: ships but is invisible, per-directory, an |
| anthropics/claude-code#83341 | — | 1 | — | — | | [BUG] Claude Code hangs on all subprocess commands when runn |
| anthropics/claude-code#82360 | — | 1 | — | — | | One-shot CronCreate silently accepts past times — rolls to n |
| anthropics/claude-code#30675 | — | 1 | — | — | | Feature request: shared context between Claude Code, Claude. |
| anthropics/claude-code#90135 | — | 1 | — | — | | [BUG] Marketplace plugin re-materialization mid-session sile |
| anthropics/claude-code#95313 | — | 1 | — | — | | [FEATURE] Request: Require user confirmation before spawning |
| anthropics/claude-code#80697 | — | 1 | — | — | | [BUG] PreToolUse hook that fails to launch is treated as a d |
| anthropics/claude-code#90017 | — | 1 | — | — | | [Bug] Model ignores root-cause analysis guardrails and optim |
| anthropics/claude-code#95475 | — | 1 | — | — | | [Feature Request] Implement confidence calibration and evide |
| anthropics/claude-code#85663 | — | 1 | — | — | | [BUG] Windows: All install methods (npm/ps1/cmd/winget) fail |
| anthropics/claude-code#84081 | — | 1 | — | — | | [BUG] Auto-update silently installs a broken 500-byte stub w |
| anthropics/claude-code#84051 | — | 1 | — | — | | Interrupted auto-update leaves claude broken (missing bin sy |
| anthropics/claude-code#94004 | — | 1 | — | — | | [BUG] Permission "Allow" prompt appeared during an |
| anthropics/claude-code#74547 | — | 5 | — | — | | [BUG] Claude Code scheduled task caused ~$500 no-op polling |
| anthropics/claude-code#88058 | — | 1 | — | — | | Feature request: per-skill display description separate from |
| anthropics/claude-code#94918 | — | 1 | — | — | | [FEATURE] Add API/webhook (event) trigger for Claude Cowork |
| anthropics/claude-code#89783 | — | 1 | — | — | | [FEATURE] Programmatically spawn multiple named child sessio |
| anthropics/claude-code#85027 | — | 1 | — | — | | VS Code agent host drops user skill descriptions, so skills |
| anthropics/claude-code#93044 | — | 1 | — | — | | Project skills beyond a session-level cap return "Unkno |
| anthropics/claude-code#91413 | — | 1 | — | — | | [BUG] Individual sessions degrade 10-30x in latency while co |
| anthropics/claude-code#92753 | — | 1 | — | — | | Scheduled-task sessions intermittently freeze permanently on |
| anthropics/claude-code#92452 | — | 1 | — | — | | Dispatch start_code_task rejects second session in same non- |
| anthropics/claude-code#75757 | — | 1 | — | — | | [Bug] Subagents billed after monthly spend limit exceeded; f |
| anthropics/claude-code#83048 | — | 1 | — | — | | [SEV-1] budget.spent() reports 72x under actual consumption |
| anthropics/claude-code#78712 | — | 1 | — | — | | /model로 저장한 default 모델을 무시하고 세션 시작 시 다른 모델(Fable)로 시작되는 문제 |
| anthropics/claude-code#82766 | — | 1 | — | — | | [BUG] VS Code sidebar model badge shows Haiku while /model r |
| anthropics/claude-code#85592 | — | 1 | — | — | | CLAUDE_CODE_SUBAGENT_MODEL silently discards explicit per-ca |
| anthropics/claude-code#83920 | — | 1 | — | — | | [Bug] Agent tool model parameter ignored; subagents inherit |
| anthropics/claude-code#80456 | — | 1 | — | — | | Model repeatedly asserts untested 'I cannot X' cap |
| anthropics/claude-code#93654 | — | 1 | — | — | | [Bug] Inverted logic in absence-of-evidence search caused JS |
| anthropics/claude-code#92845 | — | 1 | — | — | | Asserts causal explanations without verifying; treats silent |
| anthropics/claude-code#94563 | — | 1 | — | — | | Scheduled task sessions hang indefinitely (isRunning stuck t |
| anthropics/claude-code#94575 | — | 1 | — | — | | Background-agent view shows the parent's advisor model |
| anthropics/claude-code#93143 | — | 1 | — | — | | [BUG] Streamable HTTP MCP tool call still times out ("T |
| anthropics/claude-code#56913 | — | 1 | — | — | | Make autonomous Claude Code actually viable: tiered Opus bra |
| anthropics/claude-code#72745 | — | 1 | — | — | | Quality regression: no cross-session context retention, repe |
| anthropics/claude-code#94527 | — | 1 | — | — | | [BUG] Most custom groups are no longer visible in the sideba |
| anthropics/claude-code#74141 | — | 1 | — | — | | [FEATURE] Organize sessions into folders/groups in the recen |
| anthropics/claude-code#91660 | — | 1 | — | — | | [Feature Request] Add writable MCP tool to assign sessions t |
| anthropics/claude-code#88951 | — | 1 | — | — | | remoteControlAtStartup silently no-ops when OAuth is unhealt |
| anthropics/claude-code#89342 | — | 1 | — | — | | [BUG] 2026-08-24 build breaks desktop-to-desktop Remote Cont |
| anthropics/claude-code#90080 | — | 1 | — | — | | Feature request: register Remote Control for all sessions at |
| anthropics/claude-code#76884 | — | 1 | — | — | | Repeated "verified" fixes fail because verificatio |
| anthropics/claude-code#92953 | — | 1 | — | — | | [Feature Request] Implement agent spawn validation to reduce |
| anthropics/claude-code#89596 | — | 1 | — | — | | [BUG] scheduled-task sub-agents self-invoke a skill and spaw |
| anthropics/claude-code#91767 | — | 1 | — | — | | [FEATURE] Add model, stop_reason, token usage and tool timin |
| anthropics/claude-code#86724 | — | 1 | — | — | | Nested-Agent orchestrator (e.g. built-in /code-review) repor |
| anthropics/claude-code#14882 | — | 1 | — | — | | Skills consume full token count at startup instead of progre |
| anthropics/claude-code#80469 | — | 1 | — | — | | [Bug] Skill description parser incorrectly interprets angle |
| anthropics/claude-code#93324 | — | 1 | — | — | | [Bug] Terminal UI displays incorrect model name when subagen |
| anthropics/claude-code#93646 | — | 1 | — | — | | [BUG] Model shorthand `--model sonnet` resolves to Sonnet 4. |
| anthropics/claude-code#91898 | — | 1 | — | — | 👍1 | [BUG] Claude Desktop MCP: misleading "server may be unr |
| anthropics/claude-code#77058 | — | 1 | — | — | | Repeated failure to self-apply established safeguards within |
| anthropics/claude-code#53223 | — | 1 | — | — | | [BUG][SECURITY] CLAUDE.md/AGENTS.md instruction compliance i |
| anthropics/claude-code#83253 | — | 1 | — | — | | [BUG] Scheduler silently stops firing all scheduled tasks af |
| anthropics/claude-code#81241 | — | 1 | — | — | | Scheduled one-time tasks fire early / duplicate-execute, cau |
| anthropics/claude-code#89068 | — | 1 | — | — | | [Bug] Token usage discrepancy between session carryover and |
| anthropics/claude-code#93481 | — | 1 | — | — | | [Bug] Sessions start with 9% token usage already consumed |
| anthropics/claude-code#92485 | — | 1 | — | — | | [Bug] Token usage spike unexplained during idle session |
| anthropics/claude-code#85430 | — | 1 | — | — | | [BUG] PreToolUse hooks silently do not fire for one working |
| anthropics/claude-code#77832 | — | 1 | — | — | | PostCompact hook silent-miss and forced compaction detection |
| anthropics/claude-code#82565 | — | 1 | — | — | | general-purpose subagents recursively spawn: 3 requested → 2 |
| anthropics/claude-code#92621 | — | 4 | — | — | | Enable the built-in ccd_sidebar MCP tools (move_sessions, se |
| anthropics/claude-code#90822 | — | 1 | — | — | | Desktop: no supported way to bring dormant sessions back int |
| anthropics/claude-code#88738 | — | 1 | — | — | | PreToolUse hook silently stops firing mid-session and never |
| anthropics/claude-code#82416 | — | 1 | — | — | | [Bug] TypeScript LSP server wedges silently, returning empty |
| anthropics/claude-code#71609 | — | 1 | — | — | | Stale/duplicate session titles + no wrong-session safeguard |
| anthropics/claude-code#86596 | — | 1 | — | — | | "Another live session" duplicate-name check fires |
| anthropics/claude-code#92235 | — | 1 | — | — | | Mobile client creates a duplicate session on every reconnect |
| anthropics/claude-code#67636 | — | 1 | — | — | | [Bug] Parallel agent spawning causes excessive token consump |
| anthropics/claude-code#87940 | — | 1 | — | — | | Scheduled/desktop tasks stop firing once a manual session is |
| anthropics/claude-code#87356 | — | 1 | — | — | | Hooks are tool-call-scoped, not filesystem-scoped -- no way |
| anthropics/claude-code#77851 | — | 1 | — | — | | PostToolUse hooks cannot recover the original tool_input aft |
| anthropics/claude-code#88118 | — | 1 | — | — | | Read dedup does not account for PostToolUse updatedToolOutpu |
| anthropics/claude-code#92409 | — | 1 | — | — | | SendMessage tool missing in desktop Code tab session while L |
| anthropics/claude-code#71647 | — | 1 | — | — | | [BUG] VSCode sidebar "Past conversations" empty - |
| anthropics/claude-code#82728 | — | 1 | — | — | | Scheduled one-shots: 6 of 6 failed — 3 never dispatched and |
| anthropics/claude-code#91735 | — | 1 | — | — | | Non-ASCII project paths collide in ~/.claude/projects/, sile |
| anthropics/claude-code#88813 | — | 1 | — | — | | Unresolvable `@import` in CLAUDE.md fails completely silentl |
| anthropics/claude-code#91879 | — | 1 | — | — | | Scheduled-task runs leak a resident claude process per fire |
| anthropics/claude-code#91830 | — | 1 | — | — | | [FEATURE] VS Code Extension: Display persistent monthly usag |
| anthropics/claude-code#79105 | — | 1 | — | — | | [FEATURE] Living, auto-updated project narrative spanning al |
| anthropics/claude-code#86391 | — | 1 | — | — | | [BUG] Cowork scheduled tasks: WebFetch permission gate (PROV |
| anthropics/claude-code#90912 | — | 1 | — | — | | [BUG] Desktop: send_message rejects a normal interactive ses |
| anthropics/claude-code#85820 | — | 2 | — | — | | Permission requests in unattended scheduled-task sessions ha |
| anthropics/claude-code#91017 | — | 1 | — | — | | Desktop app sidebar shows no CLI sessions — index stuck sinc |
| anthropics/claude-code#84196 | — | 1 | — | — | | [FEATURE] Scheduled-task tools expose no link between a task |
| anthropics/claude-code#73158 | — | 1 | — | — | | [BUG] UTF-8 BOM in agent .md silently prevents agent registr |
| anthropics/claude-code#81306 | — | 1 | — | — | | Windows: Desktop crash wedged the MSIX package; recovery req |
| anthropics/claude-code#88308 | — | 1 | — | — | | [BUG] Scheduled-task MCP tools (list_scheduled_tasks / updat |
| anthropics/claude-code#86115 | — | 1 | — | — | | Desktop app: paused (disabled) scheduled tasks disappear fro |
| anthropics/claude-code#85804 | — | 1 | — | — | | Desktop 앱 업데이트 후 계정 로컬 상태(세션 인덱스 + 예약 루틴 + Chrome 확장 페어링) 전부 |
| anthropics/claude-code#85565 | — | 1 | — | — | | [BUG] Desktop app update silently wiped the internal schedul |
| anthropics/claude-code#33027 | — | 1 | — | — | | Scheduled tasks: 'Always allow' option missing fro |
| anthropics/claude-plugins-official#5905 | — | 1 | — | — | | telegram v0.0.7: `claude mcp list` health-check probe SIGTER |
| aryaminus/controlkeel#145 | — | 1 | — | — | | Local stdio MCP bypasses the entire tool-authorization chain |
| asciimoo/hister#626 | — | 1 | — | — | | command to import browser bookmarks |
| athola/ragentop#24 | — | 1 | — | — | | Add kill switch: pause/terminate runaway agent sessions |
| attevon-llc/OpenTranscribe#366 | — | 1 | — | — | | feat(asr): NVIDIA NeMo (Parakeet / Canary) as a first-class |
| azrtydxb/procoder#255 | — | 1 | — | — | | internal/store: TestBreakingAStaleLockIsSerialised is load-d |
| basicmachines-co/basic-memory#672 | — | 1 | — | — | | Write Basic Memory provider for Supermemory's memoryben |
| beatzball/roost#100 | — | 1 | — | — | | Shared notes and an advisory lock for a fleet, as plain file |
| beebls/youtube-music-history-scrobbler#10 | — | 1 | — | — | | takeout YouTube/YTM doesn't get all history, use MyActi |
| bengous/bookmarker#30 | — | 1 | — | — | | Firefox writer missing origin_id handling in moz_places |
| bmad-code-org/BMAD-METHOD#2725 | — | 1 | — | — | | [6.11.0] sprint_plan.py / sprint_status.py: BOM'd epic |
| browser-use/browser-use#5786 | — | 1 | — | — | | fix(history): preserve usage stats during save_to_file and m |
| browserbase/stagehand#2527 | — | 1 | — | — | | fix(locator): serialize click input events |
| browserbase/stagehand#1250 | — | 1 | — | — | | Storage state persistence broken in Stagehand v3 - userDataD |
| chigwell/telegram-mcp#150 | — | 1 | — | — | | Fix 'charmap' codec can't encode error on Win |
| chigwell/telegram-mcp#180 | — | 2 | — | — | | FloodWait is invisible in the tool runtime: short waits abso |
| code-yeongyu/oh-my-openagent#8579 | — | 1 | — | — | | spawn admission reply lost -> parent reports start_failed |
| cohere-ai/cohere-python#797 | — | 1 | — | — | | fix(utils): preserve embedding types across batches |
| dimoschi/touchstone#47 | — | 1 | — | — | | Make the quality skill portable across Codex and Claude Code |
| djm204/frankenbeast#4481 | — | 1 | — | — | | bug(observer): SpanLifecycle.recordTokenUsage diverges Token |
| drabaioli/cdd#78 | — | 1 | — | — | | Cap prose length the way the roadmap cap does: inventory and |
| elie222/inbox-zero#1880 | — | 1 | — | — | | GMail API / Rate Limit Issues |
| eltmon/overdeck#1868 | — | 1 | — | — | | Cost-bleed circuit breaker: progress-aware, always-on guard |
| evalstate/fast-agent#903 | — | 2 | — | — | | Marketplace entry name is used as an install directory name |
| fitchmultz/pi-cursor-sdk#173 | — | 1 | — | — | | Shell tool bypasses Cursor preToolUse hooks in pi sessions ( |
| fraction-owl/transit_planning_with_python#157 | — | 1 | — | — | | `SystemExit: 0` on a successful run reads as a crash to note |
| frederick-douglas-pearce/claude-code-sessions#91 | git-s7 | 1 | — | — | | reference: document conversation continuity / compaction rec |
| get-bb/bb#2328 | — | 1 | — | — | | Automatic recovery storms a dead Codex thread with repeated |
| get-bb/bb#3059 | — | 1 | — | — | | In-app browser opens signed-in services on their login form |
| get-bb/bb#2692 | — | 1 | — | — | | A script automation cannot report UNKNOWN, so honest blindne |
| ggml-org/whisper.cpp#1724 | — | 1 | — | — | | Hallucination on silence |
| google-gemini/gemini-cli#28912 | — | 1 | — | — | | You do not have a valid license of this product. |
| google/adk-python#7012 | — | 1 | — | — | | fix(tools): check_require_confirmation fails closed on non-b |
| gourabanandad/error-translator-cli-v2#99 | — | 1 | — | — | | Auto-hook should not intercept KeyboardInterrupt and SystemE |
| grammy-jiang/socratic-method#33 | — | 1 | — | — | | setup codex default (per-file symlinks) is invisible to Code |
| headroomlabs-ai/headroom#3418 | — | 1 | — | — | | [BUG] Windows persistent install is impossible without admin |
| huggingface/smolagents#2678 | — | 1 | — | — | | fix(types): fix AgentImage numpy/tensor conversion and color |
| jszmajda/lid#73 | — | 1 | — | — | | Vacuous tests read as coverage: the test level admits one in |
| jwalin-shah/inbox#69 | — | 1 | — | — | | Make Gmail incremental sync quota-safe and prevent one accou |
| karlicoss/HPI#406 | — | 1 | — | — | | Feature Request: Deduplication of Data |
| karlicoss/HPI#103 | — | 1 | — | — | | Think about abandoning timezone abbreviations map in my.core |
| karpathy/nanochat#800 | — | 1 | — | — | | [WIP] The experiment refactor: nanochat as a miniseries rese |
| kyegomez/swarms#1982 | — | 1 | — | — | | [FEAT] Enforce a read-before-write invariant on file edits |
| kytlogia/safario-historilo#180 | — | 1 | — | — | | [Bug] Chrome起動中の履歴自動読み込み失敗時に「not an error」という無意味なエラーメッセージが表示 |
| langchain-ai/langgraph#8919 | — | 1 | — | — | | checkpoint-conformance suite has no operator-form filter cov |
| langchain-ai/langgraph#8850 | — | 1 | — | — | | Root Makefile aggregate targets can report success after a c |
| langchain-ai/langgraph#8340 | — | 1 | — | — | | InMemoryStore upsert overwrites created_at on existing keys |
| langfuse/langfuse#16056 | — | 1 | — | — | | bug(annotation-queues): re-adding objects creates duplicate |
| langfuse/langfuse#15734 | — | 1 | — | — | | bug(models): model-name uniqueness relies on sentinel column |
| linuxfoundation/lfx-self-serve#2032 | — | 1 | — | — | | BUG: Sign Date in My CLAs shows UTC date, off by one day for |
| mastra-ai/mastra#24611 | — | 1 | — | — | | Scheduler: year-pinned (7-part) cron is accepted but never f |
| mem0ai/mem0#4956 | — | 1 | — | — | | ADD-only extraction in v3 may surface stale/contradictory fa |
| mem0ai/mem0#4884 | — | 1 | — | — | | BM25 keyword search and entity extraction are hardcoded to E |
| mem0ai/mem0#6933 | — | 1 | — | — | | bug(sdk-python): MilvusVectorStore returns raw distance as s |
| microsoft/semantic-kernel#13563 | — | 1 | — | — | | .Net Bug: Method IsKeyPropertyTypeValid in type Microsoft.Se |
| mila-iqia/skills#5 | — | 1 | — | — | | Switch skill CI to use docs URL instead of local SKILL.md fo |
| missingbulb/Claudinite#170 | — | 1 | — | — | | Revisit: Workflow tool's interactive opt-in blocks unat |
| mistralai/mistral-vibe#1055 | — | 1 | — | — | 👍2 | bug: MISTRAL_API_KEY silently overrides subscription-linked |
| modelcontextprotocol/go-sdk#1131 | git-s5 | 1 | — | — | | MemoryEventStore.After panics (slice out of range) on an ind |
| modelcontextprotocol/go-sdk#1105 | git-s9 | 1 | — | — | | internal/jsonrpc2: wrap writeErr with %w so errors.Is works |
| modelcontextprotocol/python-sdk#3408 | — | 1 | — | — | | test_safe_join_rejects_symlink_escape fails on Windows witho |
| monk-io/monk-plugin#496 | — | 1 | — | — | | fix(windows): write mcp_config.json as UTF-8 without BOM on |
| navotvolkgroundup/browser-migrate#3 | — | 1 | — | — | | Non-Chromium bookmark write (Firefox places.sqlite, Safari p |
| nimbalyst/nimbalyst#1489 | — | 1 | — | — | | Claude Agent sessions never see MCP servers (or hooks) shipp |
| obra/superpowers#2230 | — | 1 | — | — | | skills/test-driven-development: three gaps — tests as the de |
| obra/superpowers#931 | — | 1 | — | — | | Feature: Plan-aware session handoff commands (/create_handof |
| ohmyviv/longread-collector#159 | — | 1 | — | — | | Reliability blocker: Sheets read amplification can exceed 60 |
| omegaup/omegaup#10136 | — | 1 | — | — | | bug: SystemExit(0) in with-body recorded as failure |
| openai/codex#39223 | — | 1 | — | — | | Codex Desktop loads global AGENTS.md but stale memory overri |
| openai/codex#38347 | — | 1 | — | — | | Windows elevated sandbox fails on session-scoped mapped-driv |
| openai/codex#42080 | — | 1 | — | — | | Codex long-task reliability: completion claims without targe |
| openai/codex#45306 | — | 1 | — | — | | Codex incorrectly reports an MCP workflow as impossible with |
| openai/codex#37226 | — | 1 | — | — | | Automatically isolate and coordinate concurrent writes acros |
| openai/codex#28903 | — | 1 | — | — | | AGENTS.md not loaded from ancestor directories above repo ro |
| openai/codex#36631 | — | 1 | — | — | | PowerShell uses C:\ instead of the workspace directory for D |
| openai/openai-agents-js#1098 | — | 3 | — | — | | feat: #1097 support overrideArguments for approved tool call |
| openclaw/openclaw#128395 | — | 1 | — | — | | claude-cli backend: per-agent tools.deny never reaches --dis |
| openclaw/openclaw#84393 | — | 1 | — | — | | OpenClaw Codex runtime silently injects coding-agent base pr |
| plusky/bugwarden#203 | — | 1 | — | — | | Automated mutation testing, non-blocking only — to catch lap |
| pydantic/pydantic-ai#4773 | — | 1 | — | — | | RFC: Pluggable cross-run memory layer (AbstractMemoryStore) |
| pydantic/pydantic-ai#6460 | — | 1 | — | — | | Cancellation semantics: a level-triggered contract for agent |
| pydantic/pydantic-ai#8368 | git-s2 | 1 | — | — | | Carrying `usage=` across runs makes the agent-run span repor |
| pydantic/pydantic-ai#8023 | — | 1 | — | — | | Wire-contract cases assert against their own cassette, so th |
| pydantic/pydantic-ai#7247 | — | 1 | — | — | | First-class checkpoint/rewind primitive for agent runs |
| pyranthus-hq/mora#515 | — | 1 | — | — | | Gmail incremental sync exhausts 6,000 query-cost units with |
| rafcabezas/arc2zen#20 | — | 1 | — | — | | Add Arc browsing history migration |
| rbaumier/comply#8176 | — | 1 | — | — | | Discovery: flat-config `ignores` negations are lost to direc |
| rjmurillo/ai-agents#4880 | — | 1 | — | — | | chore(context): measure and slim path-local AGENTS.md layers |
| rlespinasse/drawio-exporter#136 | — | 1 | — | — | | read_file() does not strip a UTF-8 BOM before parsing drawio |
| roxspring/markflow#22 | — | 1 | — | — | | Lint rule: one sentence per line |
| run-llama/llama_index#23144 | — | 1 | — | — | | [Bug]: Memory silently evicts the system message from chat h |
| ruvnet/ruflo#2895 | — | 1 | — | — | | backupMemoryDb() never integrity-checks the source — corrupt |
| rysweet/azlin#1159 | — | 1 | — | — | | Provision OOM containment on hosts: a runaway agent tree kil |
| screenpipe/screenpipe#6698 | — | 1 | — | — | | [feature] private infinite memory with on-demand AI access |
| simstudioai/sim#3037 | — | 1 | — | — | | [FEATURE REQUEST] Configurable Embedding for Knowledge Base |
| sodiumsun/agenttrail#1 | — | 1 | — | — | | Recursive fs.watch exhausts the inotify limit and hard-crash |
| sourcefrog/cargo-mutants#643 | — | 1 | — | — | 👍1 | Detect nondeterministic tests before a mutation run |
| stevehansen/eidet#80 | — | 1 | — | — | | Provenance residuals from #34: unknown ≠ untrusted, and line |
| su-kaka/gcli2api#401 | — | 1 | — | — | | [Bug]: gcli2api 调用403 |
| synapt-dev/recall#854 | — | 1 | — | — | | PreCompact hook 'Hook cancelled' silently fails on |
| tenequm/pond#239 | git-s4 | 1 | — | — | | schedule: the generated systemd user timer stops re-arming, |
| topoteretes/cognee#4811 | — | 1 | — | — | | Hackathon [Feature]: Add Reddit data-source connector |
| topoteretes/cognee#4730 | — | 1 | — | — | | Hackathon [Feature]: Add Telegram data-source connector |
| topoteretes/cognee#4725 | — | 2 | — | — | | Hackathon [Feature]: Add Obsidian vault data-source connecto |
| traceloop/openllmetry#4426 | — | 1 | — | — | | 🐛 Bug Report: [OpenAI] Non-ASCII characters are escaped in ` |
| traceloop/openllmetry#4431 | — | 1 | — | — | | 🐛 Bug Report: NotGiven sentinel leak in Anthropic instrument |
| truera/trulens#1749 | — | 1 | — | — | | [BUG] String type error in trulens dashboard using `add_data |
| twentyhq/core-team-issues#1807 | — | 1 | — | — | | Date & timezone handling |
| vad1ym/oxlint-vue#2 | — | 1 | — | — | | `-c` / `--config` silently lint nothing and exit 0 |
| vitalii-zinchenko/dictara#76 | — | 1 | — | — | | [Bug]: Transcription loops/repeats same word continuously in |
| vshulcz/deja-vu#2996 | — | 1 | — | — | | Claude Code variants under ~/.cc-mirror and headless transcr |
| watt-mind/factory#1136 | — | 1 | — | — | | epic: watt-mind-factory GitHub App — own rate budget, bot id |
| xai-org/xai-sdk-python#204 | — | 1 | — | — | | fix: account polling elapsed time on the monotonic clock |
| xsmyile/sissy#126 | — | 1 | — | — | | Manage and share local skills between Claude Code and Codex |
| xwp/stream#1119 | — | 1 | — | — | | Date filtering is off by one day |
| yinggarykairui/factory-hub#90 | — | 1 | — | — | | meta: STYLE.md's '2-5 sentences' fights its o |
| yudame/research#8 | — | 1 | — | — | | Review & Update: NotebookLM Automation Plan |
| yukineko/claude-harnesses#65 | — | 1 | — | — | | schemaguard validate: enum check silently skips non-string v |
| zenml-io/zenml#5215 | — | 1 | — | — | | Idempotency support for webhook events |
| zenml-io/zenml#4682 | — | 1 | — | — | | aws-artifact-store-setup.sh breaks on us-east-1 and fails si |
| zilliztech/memsearch#703 | git-s7 | 1 | — | — | | Windows is blocked by an obsolete platform guard — milvus-li |