- Refactored SSE stream parsing to use generic `readSseJson` utility instead of domain-specific handlers across multiple packages.
- Migrated from Node.js Buffer API to Web standard APIs (Uint8Array, TextDecoder, DataView) for cross-platform compatibility.
- Simplified stream transformation pipeline by consolidating multiple stream operations into unified `createTextLineSplitter` utility.
- Refactored `ptree.ts` to remove complex stream pumping infrastructure and simplify stderr handling with direct async iteration.
- Rewrote stream utilities to operate at binary level with byte-level parsing for improved efficiency and reduced string allocations.
- Updated TypeScript configuration to include DOM.AsyncIterable type definitions for async iterable DOM API support.
- Added support for image file mentions with automatic detection, resizing, and attachment via @filepath syntax.
- Enhanced file mention messages to support both text content and image attachments with optional line count for text files.
- Implemented image dimension information display in file mention UI.
- Updated file mention processing to respect image auto-resize settings.
- Modified file mention data structure to include optional image property and optional lineCount field.
- Refactored variable naming in file processor to remove unused variable convention.
- Fixed plan mode toggles during streaming to inject plan-mode context immediately, preventing file edits while in plan mode.
- Updated file mention processing to respect image auto-resize settings.
- Added sendPlanModeContext method to inject plan mode context message into conversation history.
- Converted class properties to TypeScript parameter properties across 51 files to reduce boilerplate code.
- Removed explicit field declarations and manual assignments in constructors by using TypeScript's parameter property syntax with access modifiers.
- Applied consistent pattern of declaring private readonly and public readonly properties directly in constructor parameters.
- Modified setThinkingLevel() API to accept optional persist parameter with default value of false, allowing callers to control whether thinking level changes are persisted to settings.
- Updated SetThinkingLevelHandler type signature to include optional persist parameter for consistency across extension handlers.
- Simplified setThinkingLevel method signature from accepting options object to accepting boolean persist parameter, improving API clarity.
- Added import of expandRoleAlias function from model-resolver module.
- Added role alias expansion step before parsing model strings in model resolution logic.
- Refactored getSortedSessions to use direct array push instead of Promise.all with null filtering.
- Refactored collectSessionsFromFiles to use Promise.all for parallel file processing instead of sequential for loop.
- Simplified error handling by removing explicit null returns and filtering logic.
- Added optional `persist` parameter to `setThinkingLevel()` method to control whether thinking level changes are saved to settings.
- Changed thinking level cycling to no longer persist changes to settings by default, enabling quick iteration through thinking levels without modifying user preferences.
- Updated selector controller to handle both `thinkingLevel` and `defaultThinkingLevel` setting changes.
- Replaced nanoid and randomUUID with Snowflake ID generator across codebase for consistent distributed ID generation.
- Created new Snowflake utility module in packages/utils with support for 64-bit distributed ID generation with configurable epoch, machine ID, and sequence numbers.
- Updated session ID format documentation from nanoid to snowflake hex string format (16 hex characters).
- Removed nanoid dependency from package.json as it is now replaced by internal Snowflake implementation.
- Added 'commit' model role for dedicated commit message generation.
- Exported 'resolveModelOverride' function from model-resolver module for public use.
- Refactored model role system to centralize role configuration in MODEL_ROLES registry.
- Extracted model resolution utilities into dedicated model-resolver module for improved code organization.
- Made 'tag' and 'color' properties optional in ModelRoleInfo interface to support roles without visual indicators.
- Enhanced model selector with defensive null/undefined checks for optional role properties.
- Added default generic type parameter to Model interface, allowing Model to be used without explicit type argument.
- Removed explicit <any> generic type parameters from Model type annotations throughout codebase, leveraging new default parameter.
- Centralized model role system into MODEL_ROLES registry with metadata (tag, name, color) and MODEL_ROLE_IDS array for consistent role management across the codebase.
- Refactored model selector to dynamically generate menu actions from MODEL_ROLES registry instead of hardcoded role definitions.
- Simplified model role resolution by replacing hardcoded role checks with loops over MODEL_ROLE_IDS array.
- Removed support for 'omp/' model role prefix in favor of 'pi/' prefix throughout model selection, task execution, and configuration.
- Added getModelRoles() and overrideModelRoles() helper methods to Settings class for cleaner model role management.
- Reorganized todo-related settings from `todoCompletion.*` namespace to `todo.*` namespace with separate `todo.enabled` and `todo.reminders.*` properties.
- Updated settings access pattern in agent session from grouped settings object to flat dot-notation keys for improved consistency.
- Refactored todo_write tool availability check to independently gate on both `todo.enabled` setting and absence of submit_result.
- Fixed tool parameter schemas displaying internal TypeBox metadata fields in system prompt.
- Removed project tree snapshot generation from system prompt as unused feature.
- Removed unused imports and constants related to project tree scanning functionality.
- Added session-based shell execution with persistent shell instances and session management via sessionKey parameter.
- Added support for shell snapshots to initialize bash sessions with saved state via snapshotPath parameter.
- Added sessionEnv option to set environment variables once per session, separate from command-specific env variables.
- Added process group management functions (process_group_id and kill_process_group) for better process lifecycle control on Linux, macOS, and Windows.
- Changed env option to apply variables for the current command only, with session-level variables now managed via sessionEnv.
- Refactored shell execution to use async output queuing with enqueueChunk for proper chunk sequencing and improved timeout handling with process group termination.
- Replaced SettingsManager class with new Settings singleton providing synchronous get/set API and background persistence.
- Introduced settings-schema.ts as single source of truth for all configuration definitions with 100+ settings organized into logical groups.
- Migrated from method-based settings access (getTheme(), setTheme()) to path-based access using dot notation (settings.get('theme'), settings.set('theme', value)).
- Removed 2035-line settings-manager.ts file and replaced with modular settings.ts (693 lines) and settings-schema.ts (836 lines) for improved maintainability.
- Unified settings schema into single source of truth eliminating duplicate definitions across settings-defs.ts and settings-manager.ts.
- Added getCompactContext() API to agent session for accessing parent conversation context.
- Implemented automatic submit_result tool injection for subagents with explicit tool lists.
- Added contextFile parameter to ExecutorOptions for passing parent conversation context to subagents.
- Enhanced subagent system prompt with conditional context file instructions and improved code formatting.
- Removed schema override notification from task summary prompt to simplify output.
- Implemented formatCompactContext() method to format conversation history for subagent consumption.
- Added OpenCode Zen provider with API key-based authentication supporting multiple AI models.
- Added four new free models via OpenCode provider: glm-4.7-free, kimi-k2.5-free, minimax-m2.1-free, and trinity-large-preview-free.
- Added glm-4.7-flash model via Zai provider.
- Updated pricing and configuration for multiple models across Kimi, MiniMax, OpenRouter, Anthropic, Vercel AI Gateway, and Zai providers.
- Removed google/gemini-2.0-flash-exp:free from OpenRouter and stealth models from Vercel AI Gateway.
- Improved OAuth token refresh error handling to distinguish between definitive auth failures and transient errors.
- Added logic to only remove credentials for definitive failures (invalid_grant, revoked, expired tokens) while temporarily blocking credentials for transient errors like network timeouts.
- Enhanced error message handling in commit agent to display error messages when stopReason is 'error'.
- Added debug stack trace output for agent errors when DEBUG environment variable is set.
packages/ai:
- fix: handle "sensitive" stop reason from Anthropic API
- fix: normalize tool call IDs with special characters for Responses API
- fix: add overflow detection for Bedrock, MiniMax, Kimi providers
- fix: 429 status is rate limiting, not context overflow
packages/tui:
- fix: refactored autocomplete state tracking
- fix: file autocomplete should not trigger on empty text
- fix: configurable autocomplete max visible items
- fix: improved table column width calculation with word-aware wrapping
packages/coding-agent:
- fix: preserve external config.yml edits on save (#1046 by @nicobailonMD)
- fix: resolve macOS NFD and curly quote variants in file paths
- Added error handling for database initialization failures with detailed diagnostics including directory existence and writability checks.
- Improved directory creation in ensureDir() with proper error handling for EEXIST conditions and post-creation verification.
- Set restrictive file permissions (0o700) on created agent storage directories for improved security.
- Optimized applyLineResets() in TUI to use in-place array modification instead of creating new array via map().
- Added calculatePromptTokens() export to compute prompt-specific token counts from usage data.
- Added estimateEntriesTokens() helper to calculate total tokens across a range of session entries.
- Enhanced prepareCompaction() to dynamically adjust keepRecentTokens based on prompt-to-estimated token ratio for better context preservation.
- Improved compact() summary generation logic to preserve previous summaries when no new messages exist and handle edge cases with explicit fallback.
- Added SkillMessageComponent to render /skill command messages as compact skill entries with expandable/collapsible UI showing skill metadata and optional full prompt content.
- Added promptCustomMessage() public method to AgentSession to support custom message types with metadata and details properties.
- Added SKILL_PROMPT_MESSAGE_TYPE constant and SkillPromptDetails interface to support skill-specific message handling.
- Enhanced patch tool documentation with guidance on structured block editing, error recovery, and best practices for maintaining code integrity.
- Refactored message handling in AgentSession by extracting _promptWithMessage() private method to consolidate common logic between prompt() and promptCustomMessage().
- Added Kimi Code provider integration with OAuth device authorization flow and token management.
- Added four new Kimi Code models (kimi-for-coding, kimi-k2, kimi-k2-turbo-preview, kimi-k2.5) with reasoning support and 262K context window.
- Added kimiUsageProvider for fetching and caching Kimi Code API usage quota information.
- Added kimi-code login command to CLI for OAuth authentication with Kimi Code.
- Updated openai-completions provider to support Kimi-specific headers and cached token formats.
- Updated MiniMax-M2 model pricing: input 1.2->0.6, output 1.2->3, cacheRead 0.6->0.1.
- Extracted tool choice mapping logic into dedicated utility module for provider-specific format conversions.
- Refactored task summary generation to use template-based rendering instead of string concatenation.
- Migrated system prompt generation to use Handlebars templates with structured data instead of hardcoded strings.
- Extracted JTD to TypeScript conversion logic into dedicated utility module for schema documentation.
- Consolidated prompt template rendering across multiple modules using unified renderPromptTemplate function.
- Simplified type assertions in stream module to use generic OptionsForApi type instead of provider-specific types.
- Added session compaction auto-continue feature that automatically sends a synthetic continuation prompt after compaction completes when enabled in settings.
- Added tool output pruning system that intelligently removes verbose tool outputs from session history during compaction to reduce token usage while preserving critical information.
- Added short summary generation for compaction results, providing concise PR-style summaries of conversation changes alongside full summaries.
- Added session.compacting extension hook that allows extensions to customize compaction prompts, add context, and preserve metadata during summarization.
- Added synthetic message support to mark system-injected messages (like auto-continue prompts) and prevent them from being persisted to editor history.
- Added ToolChoice type and toolChoice parameter support across all AI providers (OpenAI, Azure OpenAI, Anthropic, Google) enabling fine-grained control over tool/function selection during LLM calls.
- Added toolChoice override capability to Agent.prompt() method and session prompt options allowing callers to control tool selection behavior per request.
- Added provider-specific tool choice mapping functions (mapAnthropicToolChoice, mapGoogleToolChoice, mapOpenAiToolChoice) to normalize tool choice formats across different LLM APIs.
- Removed kernel heartbeat/ping mechanism from PythonKernel, simplifying health monitoring by relying on direct isAlive() checks instead of periodic HTTP requests.
Creates a new session file with the exact same state as the current session,
including all entries (messages, model changes, compactions, etc.) and
artifacts. The new session's parentSession field links to the original.
- Added fork() method to SessionManager that creates new session file with copied entries
- Added fork() method to AgentSession that handles artifact directory copying
- Added /fork command handler in input-controller and command-controller
- Extended session switch event reason type to include 'fork'
- Replaced external JSON/JSONL parsing libraries with Bun's built-in JSON5 and JSONL APIs across all packages.
- Removed dependencies: json5, ndjson, get-east-asian-width, json-stringify-safe, split2, through2, and @types/ndjson.
- Replaced custom text width and ANSI wrapping implementations with Bun.stringWidth() and Bun.wrapAnsi() APIs.
- Updated Bun type definitions from ^1.2.17 to ^1.2.18 and bun-types from ^1.3.5 to ^1.3.7.
- Refactored JSONL parsing throughout codebase to use Bun.JSONL.parse() and Bun.JSONL.parseChunk() with improved buffer management and error handling.
- Updated test assertions to use Bun.stringWidth() for dynamic width calculations instead of hardcoded values.
- Inline preloaded skill contents in system prompts and allow task-level skill selection.
- Record session init entries with system prompt, task, tools, and output schema.
- Fix session persistence to keep pre-assistant entries in subagent logs.
- Added model_usage table to persist most-recently-used model timestamps across sessions.
- Model selector now sorts fuzzy-matched results by usage recency.
- Schema version bumped to 3 for model_usage table.
- Removed unused currentModel field from ModelSelectorComponent.
- Plan mode provides structured workflow where agents propose plans for user approval before execution.
- Added plan:// internal URL protocol for accessing plan files and injecting plan-mode context into subagent prompts.
- Added plan mode toggle shortcut and paused status indicator in status line.
- Fixed plan reference injection to properly pass workflow state to plan-mode system prompts.
- Improved autocomplete fuzzy matching to support subsequence matching for skill suggestions.
- Removed Prettier configuration files (.prettierignore and .prettierrc) and migrated formatting to Biome.
- Updated Biome configuration from version 2.3.11 to 2.3.12 and changed arrowParentheses rule from 'always' to 'asNeeded'.
- Pinned @biomejs/biome dependency to exact version 2.3.12 in package.json and bun.lock.
- Applied consistent arrow function formatting across 489 files by removing unnecessary parentheses around single parameters.
- Removed blank lines after comment blocks and reorganized imports for consistency across the codebase.
- Changed loadCustomModels from async to sync using fs.readFileSync for immediate resolution.
- Updated discoverModels to return synchronously, removing await from main.ts, pipeline.ts, and commit index.ts.
- Migrated sdk.ts to use node:fs instead of node:fs/promises for synchronous operations.
- Refactored scanDirectoryForSkills to use recursive traversal instead of batch Promise.all.
- Added tarball Dockerfile for testing npm publish/install flow with verdaccio registry.
- Converted readdirSync, readFileSync, and statSync to async readdir, readFile, stat across skills and agent discovery.
- Made scanDirectoryForSkills async and refactored custom directory scanning to use Promise.all for concurrent processing.
- Updated agent discovery to use fs/promises for async file reading and refactored helper patterns.
- Added AgentParsingError exception class for better error handling during agent parsing.
- Added filesystem error type guards (isEnoent, isEacces, isPerm, etc.) to pi-utils for safe error checking.
- Added color manipulation utilities to pi-utils for accessibility features.
- Added color-blind mode setting to settings manager.
- Migrated plugins, settings, and config modules from sync to async file operations.
- Updated error handling to use new pi-utils type guards for type-safe checking.
- Removed WASM generation script; use Bun `wasm?raw` loader for imports.
- Added bunfig.toml with loaders for `.md`, `.py`, and `.wasm?raw` text imports.
- Added types/assets/index.d.ts for global TypeScript module declarations.
- Unified TypeScript configuration with tsgo-based checking across monorepo.
- Removed build and WASM steps from install and publish pipelines.
- Added tsconfig.publish.json files to all packages with optimized publish-time configuration.
- Updated all package.json scripts with prepublishOnly hooks for correct type checking during publish.
- Added @oh-my-pi/omp-stats path mappings to root tsconfig.json for consistent imports.
- Added WASM generation script for photon module and integrated into install:dev script.
- Fixed indentation and line breaks in nested code blocks (lsp/render.ts, patch/shared.ts).
- Collapsed overlong function arguments and ternary expressions to standard line length.
- Reorganized import statements for consistency (ask.ts).
- Improved readability of conditional expressions (output-block.ts, web/search/render.ts, python.ts).
- Added retry utility functions `isRetryableError` and `extractHttpStatusFromError` to the @ai package for identifying transient errors suitable for retry operations.
- Made `AgentStorage.open()` asynchronous with exponential backoff retry logic to handle SQLITE_BUSY errors during concurrent database access.
- Refactored `AuthStorage` to use async factory pattern with `create()` method instead of synchronous constructor for proper async initialization.
- Added graceful error handling in terminal operations to exit cleanly when terminal becomes unavailable due to EIO errors instead of crashing.
- Added error handling during emergency terminal restore to prevent errors when terminal is already dead during crash cleanup.
- Settings now stored in ~/.omp/agent/config.yaml instead of agent.db
- Models config now supports models.yaml (preferred) alongside models.json
- Auto-migration from agent.db settings and models.json on first run
- AgentStorage.saveSettings() deprecated (getSettings() kept for migration)
- storage-migration.ts now only handles auth.json migration
- Added AgentOutputManager to allocate unique IDs for agent outputs across task invocations, preventing artifact collisions.
- Implemented automatic ID scanning in ArtifactManager to resume sessions without overwriting existing artifacts.
- Integrated AgentOutputManager into task processing to allocate unique IDs before template rendering.
- Added agentOutputManager property to ToolSession interface for managing unique agent:// IDs across task invocations.
- Fixed file descriptor leak in OutputSink when write fails
- Fixed isError flag not propagating to UI for grouped tool calls
- Fixed Python streaming updates during cell execution (missing onChunk)
- Fixed ls icon detection broken by age suffix in entry names
- Added Tool Output Infrastructure section to DEVELOPMENT.md
can1357