can1357/oh-my-pi v18.0.4

on GitHub

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

Changed

• Improved performance in append-only context mode by memoizing message serialization, keeping per-call sync overhead flat as conversations grow.

Fixed

• Fixed an issue where onTurnEnd was skipped for turns ended by a terminal tool result (such as a subagent's final yield).

@oh-my-pi/pi-ai

Fixed

• Fixed Cursor tool calls through OpenAI-compatible authentication gateways losing arguments when complete argument maps are sent without streaming deltas (#9479).
• Fixed Cursor plan entitlement refusals repeatedly selecting ineligible accounts by scoping credential blocks to the requested model during rotation (#9488).
• Improved HTTP 413 error classification to accurately distinguish between payload/media size limits and token context window overflows, preventing inappropriate token compaction attempts and routing to correct recovery/fallback strategies (#9235).
• Fixed Cursor conversation rotation after aborts or mid-turn restarts to properly replay the last user message on a fresh conversation.

@oh-my-pi/pi-catalog

Fixed

• Fixed default reasoning effort for cursor/cursor-grok-4.5 and cursor/cursor-grok-4.6 so requests without an explicit effort setting default to -medium instead of -low, preventing rate limit rejections on Cursor's Start plan (#9478).
• Fixed aliased OpenCode Zen Ox Alpha models exposing incorrect effort levels, ensuring the gateway's native low, high, and max tiers are correctly mapped and reachable (#9349).
• Fixed missing rate card tier for public xAI and SuperGrok models with prompt lengths exceeding 200K tokens (#9512).

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

Added

• Added the omp git command (and /git slash command): an interactive, fullscreen repository TUI featuring a split/inline/hunk diff viewer with minimap scrollbar, syntax highlighting, a staging sidebar with line-level staging, commit composer with amend support, and author avatars. Supports keyboard navigation, full mouse interaction, and pinning views to specific commits via omp git <revision>.
• Overhauled the /extensions Extension Control Center into a fullscreen alternate-screen dashboard with mouse support, tab navigation, unified inspector views across extension types, live MCP connection management, and expandable details (Ctrl+O).
• Added support for live syntax highlighting in streaming markdown code blocks.
• Added an immediately editable startup composer for interactive launches, preserving drafts typed while session initialization is in progress.

Changed

• Improved streaming markdown and thinking block rendering performance on long sessions by batching token updates and eliminating redundant re-processing.
• Optimized streaming edit verification and session restoration for large files and history-heavy sessions.

Fixed

• Fixed invalid streamed edit patches occasionally reaching the edit tool instead of being stopped early.
• Fixed ! shell commands on zsh/fish by running them inside a real PTY, resolving terminal option errors and preserving ANSI color formatting.
• Fixed transcript layout corruption and viewport compression caused by interrupted streams, empty blocks, or collapsed wrapped diff lines.
• Fixed transcript scrollback loss where output below sticky cards (such as hub-wait or todo) failed to commit to terminal history.
• Improved HTTP 413 error handling: accurately distinguish between true token-context overflows and provider byte/media budget limits, persist terminal errors across sessions, and enable proper fallback-chain model switching.
• Fixed discovery-backed session models failing to restore when resuming sessions with omp --resume or --continue.
• Fixed browser tool initial launch timeouts on slow or cold host environments.
• Fixed eval runtime probes hanging on Windows due to inherited stdin handles.
• Fixed Claude models replaying partial thinking blocks as conversation text when interrupted mid-turn.
• Fixed image request failures with Kimi Code and Moonshot models by ensuring inline base64 image delivery.
• Fixed SQLite WAL-mode databases without sidecars failing to open in the Read tool.
• Fixed pasted image thumbnail rendering in the composer attachment preview.
• Fixed Linux startup event loop delays caused by legacy extension cache fsync churn.
• Fixed subagent advisors abandoning reviews on the final yield turn during session teardown.
• Fixed /todo expand/collapse commands and corrected /shake thinking reporting.

@oh-my-pi/hashline

Changed

• Improved line-ending normalization performance by avoiding full scan-and-copy operations on files without carriage returns, eliminating stalls on large LF-only files.

@oh-my-pi/omp-stats

Fixed

• Fixed SuperGrok usage appearing as free by applying matching public xAI API pricing (including 200K-token rates), labeling costs as API-equivalent estimates, backfilling existing usage records, and displaying subscription-only models as N/A (#9512).

@oh-my-pi/pi-tui

Changed

• Significantly improved streaming Markdown rendering performance by caching unchanged rows, resuming boundary walks, and inspecting only text deltas for guard scans and OSC 8 normalization.

Fixed

• Fixed TUI aborting when syntax highlighting fails during Markdown rendering by falling back to unhighlighted text.
• Fixed Korean IME cursor drift in Orca by properly matching two-cell Hangul Compatibility Jamo rendering.

@oh-my-pi/pi-utils

Added

• Exported getAvatarCacheDir to resolve the avatar cache directory path.

What's Changed

• fix(eval): escalate unconfirmed kernel shutdown to a process-group kill by @Mustaqeem66 in #9222
• fix(prompts): scope the delete-safety rule so clean cutover is not self-contradictory by @Mustaqeem66 in #9241
• feat(coding-agent): overhaul the /extensions inspector by @atacolak in #9311
• fix(hindsight): preserve conversation timestamps by @atacolak in #9366
• perf(agent): memoize AppendOnlyContextManager message digests by object identity by @LukasParke in #9386
• fix(tui): recognize Orca's Hangul jamo width by @roboomp in #9398
• fix(tui): do not crash streaming fences when HighlightStream is missing by @dylanjkl in #9404
• fix(agent): route payload-size 413s away from token-context compaction by @ConfoundingVariables in #9406
• fix(cursor): make #8345 conversation rotation actually recover retry/resume turns by @dperhar in #9412
• fix(catalog): normalized aliased OpenCode Zen ox-alpha effort ladders by @coderredlab in #9435
• fix(eval): bound runtime availability probes with timeout and stdin ignore by @roboomp in #9467
• perf(coding-agent): coalesce caught-up streaming reveal renders to the reveal cadence by @szavadsky in #9476
• fix(ai): preserve Cursor gateway tool arguments by @roboomp in #9481
• fix(catalog): default effort-less cursor grok to the medium tier by @roboomp in #9484
• fix(tui): drop empty transcript blocks from viewport composition by @roboomp in #9485
• fix(ai): scope Cursor plan refusals by model by @roboomp in #9489
• fix(tui): wire todo HUD expansion commands by @roboomp in #9495
• fix(tui): preserve retired welcome rows across resize by @pedropaulovc in #9496
• test(coding-agent): isolate WSL host context discovery by @pedropaulovc in #9497
• test(tui): cover physical Option forward delete by @shawnkoh in #9501
• fix(coding-agent): restore discovery-backed models on session resume by @roboomp in #9504
• fix(agent): review subagent final yield in advisor by @roboomp in #9506
• fix(acp): stop dumping structured tool progress as JSON content by @roboomp in #9514
• fix(stats): estimate public xai oauth costs by @roboomp in #9515
• fix(coding-agent): preserve subagent output artifact on hub revival by @roboomp in #9519
• fix(tui): repaint plugins settings tab after async list mounts by @roboomp in #9528
• fix(extensions): use WAL for legacy pi parse cache by @roboomp in #9551
• docs: replaced unavailable README captures by @roboomp in #9472

New Contributors

• @LukasParke made their first contribution in #9386
• @dperhar made their first contribution in #9412
• @coderredlab made their first contribution in #9435

Full Changelog: v18.0.3...v18.0.4