GitHub Release: can1357/oh-my-pi v18.4.0 (28.09.2026)

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

Changed

• Updated telemetry attribute names from the pi.* namespace to the omp.* namespace.

@oh-my-pi/pi-ai

Breaking Changes

• Changed LimitsApi.rotate() to return a CredentialRotation object ({ switched, afterSiblingWait? }) instead of a boolean. Check .switched explicitly, since the returned object is always truthy.

Added

• Added image metadata to hosted OpenAI image-generation results, including the model used and each generated image’s dimensions and quality.

Changed

• Redesigned the browser page displayed during OAuth login.

Fixed

• Fixed Anthropic accounts remaining blocked after quota resets, including for broker-connected clients, while preserving independent authentication and model-limit restrictions.


• Fixed concurrent usage refreshes repeatedly probing providers after failures, reducing unnecessary usage-endpoint requests.


• Fixed Gemini and Antigravity usage and cost reporting when upstream responses omit prompt-token counts or report more cached tokens than prompt tokens.


• Fixed credential failover after authentication errors so stored credentials rotate through every distinct sibling instead of stopping after one attempt; rotation can also wait briefly for a temporarily blocked healthy sibling to become available.


• Fixed native judge responses without token counts producing invalid usage and cost values.


• Fixed Cursor usage, cost, and context accounting to include final input, cache, reasoning, and output metrics, improving compaction and handoff sizing.


• Fixed Cursor MCP tool calls routed through external executors being returned as text instead of tool calls.


• Fixed Cursor shell-tool timeouts being displayed in milliseconds rather than seconds.


• Fixed fetch requests over Unix sockets when PI_PROXY is configured.


• Fixed Ollama chat turns being recorded with zero cost; usage is now priced using the model’s cost information.


• Fixed Anthropic requests failing after native compaction when per-message effort settings were present; effort controls are now handled correctly with compaction.

@oh-my-pi/pi-catalog

Fixed

• Corrected Cerebras model pricing so cached input tokens are charged at the model’s input rate, consistent with Cerebras billing.

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

Added

• Added the telemetry.otlpExportEnabled setting under Settings → Providers → Privacy to disable OTLP trace, log, and metric export even when OTEL_* endpoints are configured; exporting remains enabled by default.


• Added a first-launch warning when Python evaluation is enabled but no working Python interpreter is available, with guidance for configuring python.interpreter and checking the installation with omp setup python --check.

Changed

• Updated omp stats and /stats to open the redesigned dashboard immediately while session data synchronizes in the background with live progress; --json and --summary continue to synchronize before producing output.


• Replaced the stats dashboard’s Behavior page with a Frustration page that can classify messages using the judge model role, showing an estimated cost before analysis and recording /stats spending in the current session.


• Clarified the eval tool documentation to explain that its kernel may be shared with the parent session and concurrent task subagents.

Fixed

• Fixed /tree reopening saved Ask results instead of navigating past them when an optional preview was saved as null.


• Fixed the legacy createGrepTool() API when searching with both a file path and a glob filter.


• Improved task and subagent reliability: eligible saved usage resets are now redeemed automatically when polling is throttled or transient failures occur, concurrent tasks share confirmed resets, headless subagents retain assignments across session transitions, tagged ^model agents are available to nested subagents, and wait returns promptly with information about still-running work when no owned jobs are available.


• Fixed SDK requests using ApiKeyResolver to wait for a nearby healthy credential when a drained account’s quota block is about to expire, instead of immediately failing with a multi-hour quota error.


• Fixed Anthropic requests failing after native compaction when experimental context notes were enabled.


• Fixed Windows path handling for 8.3 short paths, including project-directory detection and home-directory display in status, tool labels, and errors.


• Fixed edit PUT >N producing syntactically invalid code when inserting shallower constructs near closing braces.


• Fixed Windows one-shot commands, including omp update, incorrectly reporting successful completion as an error; also fixed this behavior when no user npm or Bun configuration file exists.


• Fixed missing judge token counts corrupting session usage totals and displaying $NaN.


• Fixed Cursor sessions under-reporting token usage and cost, compacting based on the wrong context measurement, and applying shell-command timeouts in the wrong units.


• Updated goal mode to wait for user input when all remaining todos are blocked instead of repeatedly requesting approval.


• Fixed pi-background-tasks 2.6.0 and later failing to load due to a missing legacy pi-ai compatibility export.


• Fixed blob broker requests when PI_PROXY is configured.


• Fixed generate_image reporting the catalog model instead of the image model actually used by the ChatGPT/Codex backend; saved image metadata now reflects the provider-returned size and quality.


• Fixed fast-model fallback selection so it no longer chooses Gemini or MiniMax models when no smol role is configured.


• Fixed extension tool renderers using upstream pi’s renderCall(args, theme, context) signature failing to render.


• Fixed Nix flake and NixOS module builds failing because the native package version stamp was not recognized.


• Fixed Nix dependency-lock checks failing after obsolete stats chart dependencies were removed.

@oh-my-pi/collab-web

Changed

• Redesigned the web client: black chassis with one inset session panel, glass top bar with the omp mark and a live status pill, a docked composer card, prompts shown as cards in the transcript, a sectioned agents rail, and a floating agent drawer; the connect screen was rebuilt too

@oh-my-pi/pi-natives

Added

• Added Windows path utilities for converting between long and 8.3 short path spellings without resolving symlinks or junctions. Import them from @oh-my-pi/pi-natives/path.

Fixed

• Fixed the native xargs builtin so -P/--max-procs correctly limits parallel command execution, including GNU-compatible -P 0 behavior.


• Improved snapcompact rendering performance for stretched shapes on Windows x64, Intel Macs, and Linux CPUs without AVX2, with no visual changes.

@oh-my-pi/omp-stats

Changed

• Redesigned the dashboard: new layout, navigation (g + letter to jump, 1–6 to pick a range), sortable tables and in-house charts on every page


• The dashboard opens immediately and ingests sessions in the background, most recent activity first, with live progress in the header; pages update as data lands and new session activity appears within seconds


• Switching time ranges is now near-instant: range queries read hourly rollups instead of scanning every request (seconds → milliseconds on large histories)


• Frustration judge runs now scale how many requests they keep in flight (up to 256, backing off on rate limits and retrying after a short delay) and save verdicts in batches; the progress line shows the rate and requests in flight


• Provider subscription windows load separately from the rest of the Providers page, and the Projects page lists the 2,000 busiest folders


• Replaced the dashboard's Behavior page with a Frustration page showing, per model version, how often users are annoyed, annoyed at the assistant, and angry at it; a button judges unclassified messages with the host's judge model (after a cost estimate) and the chart refines live as verdicts arrive

Fixed

• Made session-statistics synchronization atomic to prevent duplicate entries when synchronization is interrupted.


• Improved the speed and reliability of initial and repeat omp stats imports, including large session histories and already-indexed histories.

@oh-my-pi/pi-tui

Fixed

• Prevented interrupted wait operations from appearing in the transcript when a queued completion or message is received.

What's Changed

• perf(pi-natives): stopped snapcompact resize lowering mul_add to libm fmaf on x86-64-v2 by @H4vC in #13496


• fix(coding-agent): read optional registry config via fs.promises in omp update by @H4vC in #13514


• fix(cli): kept exit code when loop resumed after beforeExit by @roboomp in #13472


• fix(coding-agent): wait ladder restored to old hub behavior by @H4vC in #13516


• fix(windows): normalize working directories and abbreviate short home paths by @CoderTCY in #13394


• feat(telemetry): add opt-out setting override to disable OTLP export by @krizh-p in #13444


• docs(coding-agent): say the eval kernel is shared with concurrent task subagents by @radkawar in #13521


• fix(pi-edit): keep PUT >N inserts before a closer when moving them breaks the parse by @radkawar in #13520


• feat(coding-agent): eval backend missing by @H4vC in #13529


• test: remove useless tests by @H4vC in #13539


• fix(natives): allowed const-fn lint on windows path exports by @roboomp in #13548


• fix(task): retried dropped headless prompts before yield by @roboomp in #13541


• fix(goal): stop continuations while todos are blocked by @roboomp in #13497


• docs(bash): clarified shellpath execution scope by @roboomp in #13420


• fix(judgment): handled missing judge token counts by @roboomp in #13492


• fix(nix): stamped the native addon version in the nix build by @roboomp in #13506


• fix(ai): price ollama-chat terminal usage from the model cost card by @roboomp in #13057


• fix(extensions): accept pi-order renderCall theme argument by @roboomp in #13086


• fix(coding-agent): stopped smol mini pattern matching gemini models by @roboomp in #13410


• fix(coding-agent): reported the image model the codex backend ran by @roboomp in #13404


• fix(ai): bypassed proxy for unix-socket fetches by @roboomp in #13510


• fix(coding-agent): restore anthropic compat adapter by @roboomp in #13251


• fix(ai,coding-agent): recovered cursor turn usage and tool handoffs by @roboomp in #13085


• fix(ai): waited out briefly blocked siblings before failing rotation by @roboomp in #13409


• fix(ai): restore exec-resolved marking for streamed cursor MCP calls by @H4vC in #13554


• fix(ai): drop unreachable cursor exec-resolved branch and its stale test by @H4vC in #13563


• docs(changelog): note cursor MCP exec-resolution cleanup by @H4vC in #13565


• fix(ai): rotated 401s through every distinct stored sibling by @roboomp in #13561


• fix(ai,coding-agent): keep the Anthropic compaction block first in the request by @H4vC in #13569


• fix(tui): restored saved ask questions with null previews by @roboomp in #13573


• fix(coding-agent): corrected legacy grep file glob handling by @roboomp in #13574

New Contributors

• @krizh-p made their first contribution in #13444

Full Changelog: v18.3.5...v18.4.0