github can1357/oh-my-pi v17.0.2
on GitHub

12 hours ago

@oh-my-pi/pi-agent-core

Fixed

  • Improved error visibility in interactive clients by surfacing provider stream failures through the assistant message lifecycle, preventing silent loading spinners.
  • Fixed an issue where Cursor provider contexts omitted host-supplied MCP tools from main and side-channel requests.

@oh-my-pi/pi-ai

Fixed

  • Automatically invalidate and rotate OAuth credentials when an "invalidated oauth token" error occurs.
  • Fixed auth-broker snapshot validation rejecting API keys stored via the /login flow, restoring support for gateway/broker setups serving login-sourced keys on custom hosts.
  • Fixed an issue where literal reasoning tags (e.g., <think>) inside Markdown code blocks or inline code were incorrectly treated as reasoning boundaries, which corrupted the rendered Markdown.
  • Classified HTTP 402 and "balance exhausted" quota responses as persistent usage limits, enabling automatic rotation of multi-account requests to a sibling credential.
  • Fixed kimi-code Anthropic-format requests ignoring custom provider base URLs.
  • Fixed an issue where GPT-5.6 Codex Responses-Lite requests failed with an HTTP 400 error due to invalid tool_choice parameters after tools were rewritten, by automatically downgrading forced hosted choices to tool_choice: "auto" while preserving explicit tool-use constraints.
  • Fixed Cursor streams prematurely reporting success before late CONNECT or gRPC terminal failures were observed, and resolved issues rejecting transport ends without a turnEnded signal.

@oh-my-pi/pi-catalog

Changed

  • Increased the maximum output tokens (maxTokens) from 32,768 to 65,536 for Kimi K2.7-Code models on Fireworks.

Fixed

  • Fixed a regression where the context window for openai-codex GPT-5.6 models (Luna, Sol, Terra) incorrectly fell back to 272,000 instead of preserving its 372,000 capacity.
  • Fixed Umans PAYG models incorrectly displaying as "Free" in /models by correctly sourcing their published per-token rates.
  • Fixed native moonshot/kimi-k3 capabilities and pricing, ensuring it correctly reflects its official pricing, 1M context window, image input support, reasoning capabilities, and 128k output token limit.

@oh-my-pi/pi-coding-agent

Added

  • Added native Warp CLI-agent events for rich session status, tool approvals, and completion notifications.
  • Added support for ChatGPT/Codex subscriptions in the generate_image tool, allowing image generation without a metered OPENAI_API_KEY even when using other active models.
  • Added an optional provider parameter to generate_image to override the global providers.image setting for a single request.
  • Added OpenTelemetry log and metric export support alongside existing trace exports, enabling forwarding of centralized-logger events and GenAI-semconv metrics when configured.
  • Enhanced retry.fallbackChains wildcards to support id-prefixed targets and keys, allowing more flexible model fallback routing across different providers.
  • Added an opt-in per-project model role storage mode with global fallback from the model selector.
  • Added per-advisor on/off toggle (enabled: false in WATCHDOG.yml): advisors stay in the roster but their runtime is never built — they show ○ in /advisor status rather than disappearing. Existing configs are backward-compatible (defaults to true when absent).
  • Colored the status line's advisor ++ badge by roster health (green all running, yellow quota-exhausted, red failed, dim paused); per-advisor glyphs (●/○/✕) show in /advisor status.
  • Added real provider quota display (usage percent, window, reset timer) to /advisor status and the /advisor configure preview.

Changed

  • Changed Bash command timeouts to render with a warning (yellow) border instead of an error (red) border, while still indicating to the model that the command did not complete normally.
  • Made the hashline seen-line guard opt-in and off by default via edit.enforceSeenLines, and improved handling of column-clipped lines so single-line edits on long lines apply without a full-width re-read.
  • Changed bundled TTSR rules to warn without interrupting generation.
  • Renamed the system prompt's project-context section wrapper from <context> to <repo-rules> to prevent collisions with the task tool's context parameter.
  • Enriched /advisor status to show per-advisor status glyphs, model, spend breakdown, and quota window for every configured advisor (including disabled ones), replacing the previous single-advisor-only summary.

Fixed

  • Fixed loading issues for linked legacy extensions importing DefaultPackageManager or linkedom.
  • Fixed the advisor retrying terminal, non-retriable provider failures (e.g., blocked prompts), ensuring they fail immediately while transient failures still retry.
  • Fixed an issue where reassigning the plan role model mid-planning did not take effect until the next plan-mode entry; it now applies at the next turn boundary.
  • Added managed timer helpers (ctx.setInterval, ctx.setTimeout, ctx.clearTimer) to the extension context to prevent self-scheduled callbacks from throwing uncaught exceptions and crashing the session.
  • Fixed /quit and /exit leaving stalled automatic title-generation requests alive during session teardown.
  • Fixed startup.quiet still rendering the xdev: xd://: mounted status line when MCP tools connect.
  • Fixed command errors in the hub tool when using a non-POSIX shell.
  • Fixed xdev-routed checkpoint and rewind writes not tracking checkpoint state.
  • Fixed the built-in advisor silently failing when its model routes through the Cursor provider by adding a Cursor execution bridge scoped to its granted tool set.
  • Fixed the fullscreen plan-review overlay staying visible until the approved execution turn finished; it is now hidden as soon as execution begins.
  • Fixed MCP tools repeatedly unmounting and remounting mid-session due to overlapping sanitized prefixes, and resolved stale tools remaining registered after disconnecting a server with special characters.
  • Fixed the /usage show marker and TUI usage panel to display and preserve the active organization suffix (email (OrgName)) to distinguish between multiple credentials with the same email.
  • Fixed Windows stdio MCP servers launched through .cmd or .bat shims failing with Transport closed by properly escaping arguments and spawning with windowsVerbatimArguments.
  • Fixed a startup crash on Windows when running from a drive root (e.g., R:\).
  • Fixed unknown __omp_worker_* CLI selectors exiting with code 0 instead of throwing an error.
  • Fixed Plan Review capturing mouse drags as pointer events, which prevented native terminal text selection.
  • Fixed orphaned TUI processes remaining alive after a fatal error and causing high resource consumption.
  • Fixed approved-plan execution looping through filesystem searches when a model rewrites the required plan read path.
  • Fixed Ask dialogs immediately accepting highlighted answers when they appear while the user is typing a space.
  • Stopped post-compaction auto-continue from opening another primary turn after a terminal text answer with no queued work.
  • Fixed the write approval gate misclassifying xd:// device writes as exec when the mounted tool declared a function-valued approval.
  • Fixed custom LSP servers (such as roslyn-language-server) crashing when requesting unconfigured workspace configuration sections.
  • Fixed late user-initiated bash results being recorded in whichever session or branch was active when execution finished; they now retain their originating transcript.
  • Fixed Claude Code marketplace plugins with scope: "local" leaking skills, hooks, tools, commands, and MCP servers into unrelated projects.
  • Fixed headless omp -p waiting indefinitely after a completed turn when final consolidation stalls.
  • Fixed explicit-tool sessions bypassing xd:// presentation for ambient discoverable custom and MCP tools.
  • Fixed providers.webSearch: kimi incorrectly sending Moonshot credentials instead of Kimi Code credentials.
  • Fixed registerTool demoting essential built-in tools to discoverable when a re-registration omitted loadMode.
  • Fixed the advisor skipping the next user instruction after auto-learn accepted and pruned a terminal empty assistant stop.
  • Fixed built-in advisors retrying quota- or rate-limited providers instead of applying the matching retry.fallbackChains model chain.
  • Updated model selector status messages to use role tags (SMOL, SLOW) instead of display names (Fast, Thinking) for consistency.
  • Fixed Cursor models receiving only top-level tools by forwarding mounted xd:// devices through Cursor's request-context MCP catalog.
  • Fixed Windows bash crashes when a piped command times out while flushing output.
  • Fixed a race condition where hub/IRC send and ensureLive could inject into a subagent session during disposal.
  • Migrated legacy nested/dotted configuration keys (dev.autoqa.consent and todo.reminders.max) to flat keys (dev.autoqaConsent and todo.remindersMax) on settings load.
  • Fixed keyboard input dying after the first keypress when a custom tool module attaches a stdin consumer at import time.
  • Fixed the alignment of the ask tool's custom-input dialog to start at column 0.
  • Fixed advisor context maintenance undercounting the provider context by anchoring compaction decisions on provider-reported context usage.
  • Fixed RPC mode (--mode rpc) crashing on non-JSON stdin lines; malformed lines are now reported as errors while the frame loop continues.
  • Fixed local llama.cpp Qwen-family models not honoring the --thinking off flag.
  • Documented the ultrathink, orchestrate, and workflowz magic keywords, including their effects, matching rules, and settings.
  • Fixed /clear autocomplete selecting /autoresearch and updated /clear to start a new session as an alias for /new.
  • Fixed /review aborting entirely when GitHub rejects a pull request's aggregate diff for exceeding the line limit by falling back to the paginated per-file endpoint.
  • Fixed /q + Enter running /queue instead of /quit by adding an explicit q alias to /quit.
  • Fixed Ctrl+L (app.display.reset) not refreshing the dark/light theme on certain terminals by issuing a background re-query before repainting.
  • Fixed a failing advisor stalling the primary agent: the per-turn catch-up gate parked the primary for up to its full 30s budget while a broken advisor (unsupported model, dead endpoint, render bug) retried — and an advisor exception could abort the primary's turn-end outright. A failing advisor now releases parked waiters the moment its turn fails (before any async hook), refuses new parks until a turn succeeds, and the turn-end boundary isolates advisor exceptions completely; a failed render restores the delta cursor so nothing is lost when the advisor recovers.
  • Fixed advisors retrying a permanently rejected request forever (e.g. invalid_request_error: model not supported with this account): unlike quota exhaustion — which pauses with a notice until an explicit reset — this class notified once and silently kept re-attempting every turn, re-building heavy context in a shared daemon. The runtime now hard-stops after a permanent rejection or three consecutive backlog-drop cycles, with a visible notice; an explicit reset (/new, config rebuild, restart) re-enables it. waitForCatchup resolves immediately while halted so the primary agent is never parked on a runtime that cannot drain.

@oh-my-pi/pi-natives

Fixed

  • Fixed an issue where running uv run --extra <package> pytest bypassed native pytest minimization due to a wrapper parsing error.
  • Fixed a bug where timed-out shell pipelines dropped captured output and could cause Windows hosts to terminate during teardown. (#5316)

@oh-my-pi/omp-stats

Fixed

  • Fixed the Recent Errors list to honor the selected dashboard time range before returning the newest 50 failures.

@oh-my-pi/pi-tui

Added

  • Added a fullscreen overlay mouse-tracking opt-out to allow selection-first dialogs to preserve native terminal text selection.
  • Added Terminal.refreshAppearance() to allow consumers to manually trigger a refresh of the detected dark/light terminal appearance without periodic polling.

Fixed

  • Fixed an issue where pressing Enter to accept a mid-prompt /skill:<name> autocomplete would submit and clear the draft; it now correctly inserts the skill token and leaves the prompt open.
  • Fixed Markdown rendering incorrectly turning local file paths containing www. or protocol sequences into HTTP links by requiring a valid GFM left boundary for autolinks.
  • Fixed terminal resize behavior by restoring alternate-screen rendering during drag frames, preventing wrapped fragments from polluting native scrollback while preserving the overlay-exit flicker fix.
  • Added optional right-border scrollbar to the Editor component (setScrollbarVisible): shows a thumb glyph on the right border when content overflows maxHeight, enabling scrollable multi-line editors (e.g. advisor instructions) without losing the submit hint off-screen.

@oh-my-pi/pi-utils

Added

  • Added a structured log sink API (registerLogSink, LogEvent, LogLevel) to the centralized logger, enabling out-of-band consumers (such as OpenTelemetry) to receive log events without affecting local file or console logging.

Changed

  • Bounded default ptree.ChildProcess stderr retention to a 32 KiB tail to prevent memory leaks in long-lived subprocesses. Full stderr capture must now be explicitly requested at spawn time using { stderr: "full" } on spawn or exec.

Fixed

  • Fixed fatal cleanup failing to reach process.exit() when terminal stderr is revoked.
  • Isolated rotating log files and audit state per process to prevent concurrent instances from racing during compression and rotation.

What's Changed

  • fix(agent): surface provider stream failures by @DarkPhilosophy in #5538
  • fix(tui): guard stdin against custom-tool import hijack (#5618) by @roboomp in #5744
  • fix(discovery): isolate Claude local marketplace plugins by @roboomp in #5752
  • fix(utils): bound default ptree stderr retention by @roboomp in #5761
  • fix(web-search): scope Kimi search to Kimi Code credentials by @roboomp in #5763
  • fix(codex): force tool_choice auto on responses-lite requests by @roboomp in #5772
  • feat(coding-agent): generate_image per-request provider + Codex-subscription images by @insodimension in #5321
  • fix(tui): preserve drafts on skill autocomplete by @roboomp in #5414
  • fix(rpc): tolerate malformed stdin lines instead of crashing by @roboomp in #5442
  • fix(advisor): anchor context maintenance on provider usage by @roboomp in #5463
  • fix(tui): align ask "Other" custom-input chrome to prompt gutter by @roboomp in #5475
  • fix(cli): alias /clear to /new by @roboomp in #5486
  • fix(review): fall back to per-file API when PR diff exceeds 20k lines by @roboomp in #5500
  • feat(telemetry): support full OTel — log and metric export by @roboomp in #5507
  • fix: render bash timeout with warning border instead of error by @oldschoola in #5546
  • fix(tui): use role tag in model selector status messages by @roboomp in #5586
  • docs: document magic keywords by @roboomp in #5591
  • feat(warp): emit native CLI-agent events by @metaphorics in #5592
  • feat(coding-agent): support per-project model roles by @infernix in #5596
  • fix(coding-agent): disable thinking on local llama.cpp Qwen models by @paralin in #5635
  • fix(ai): wait for Cursor protocol end after turnEnded by @roboomp in #5636
  • fix(registry): coordinate park/dispose with ensureLive and IRC delivery by @roboomp in #5637
  • fix(coding-agent): migrate legacy nested autoqa/todo settings keys by @roboomp in #5641
  • fix(cursor): expose xd devices and external MCP tools by @roboomp in #5651
  • fix(tui): constrain GFM autolinks to a valid left boundary by @roboomp in #5656
  • fix(plugins): restore legacy package manager and CJS loading by @roboomp in #5660
  • fix(plan): reapply plan role model when reassigned mid-planning by @roboomp in #5662
  • fix(extensions): isolate self-scheduled callbacks from killing the session by @roboomp in #5667
  • fix(session): abort title generation during dispose by @roboomp in #5668
  • fix(bash): drain piped output before timeout return by @roboomp in #5671
  • fix(session): honor quiet startup for xdev mount notices by @roboomp in #5672
  • fix(ai): keep literal think tags inside markdown code visible by @roboomp in #5673
  • fix(tools): fix hub tools env selection by @Parsifa1 in #5682
  • fix(session): recognize checkpoint and rewind through xdev write dispatch by @lyc-aon in #5683
  • fix(cursor): wire advisor tools through the cursor exec bridge by @roboomp in #5686
  • fix(tui): close plan review overlay before execution dispatch by @roboomp in #5689
  • fix(mcp): match tool ownership by server name, not tool-name prefix by @korri123 in #5690
  • fix(usage): org-qualified session marker for same-email accounts by @roboomp in #5692
  • fix(shell): parse uv run --extra values by @GratefulDave in #5694
  • fix(mcp): correct Windows cmd shim spawning by @roboomp in #5699
  • fix(tui): preserve organization suffixes in usage panel by @roboomp in #5703
  • fix(catalog): pinned gpt-5.6 codex context window to 372k by @roboomp in #5707
  • fix(coding-agent/launch): handle EISDIR from realpath on drive roots by @roboomp in #5709
  • fix(cli): error on unknown _omp_worker selectors by @roboomp in #5713
  • fix(tui): preserve Plan Review text selection by @roboomp in #5714
  • fix: bound fatal TUI teardown and isolate log rotation by @roboomp in #5718
  • fix(read): recover approved plans from cwd aliases by @roboomp in #5719
  • fix(tui): require Enter for single-select Ask answers by @roboomp in #5720
  • fix(coding-agent): stop post-turn maintenance turns by @roboomp in #5721
  • fix(ai): classify 402/balance-exhausted quota errors as usage limits by @metaphorics in #5723
  • fix(ai): honor custom kimi-code base URL in Anthropic format by @roboomp in #5724
  • fix(write): evaluate function-valued xd:// device approvals by @roboomp in #5729
  • fix(advisor): reconcile rewritten transcript prefixes by @roboomp in #5734
  • fix(catalog): source Umans PAYG model pricing by @roboomp in #5735
  • fix(lsp): return null for missing workspace configuration by @roboomp in #5746
  • fix(session): retain bash ownership across session/branch transitions by @roboomp in #5747
  • fix(advisor): apply configured fallback chains by @roboomp in #5748
  • fix(cli): bound print-mode mnemopi teardown by @roboomp in #5754
  • fix(catalog): correct moonshot/kimi-k3 pricing and reasoning transport by @roboomp in #5758
  • fix(tools): keep essential built-ins top-level when re-registered without loadMode by @roboomp in #5768

Full Changelog: v17.0.1...v17.0.2

Check out latest releases or
releases around can1357/oh-my-pi v17.0.2

Don't miss a new oh-my-pi release

NewReleases is sending notifications on new releases.

Get notifications