Fixed crash when rendering text containing emojis followed by long content.
The breakLongWord function was iterating over UTF-16 code units instead of
grapheme clusters, causing emojis (surrogate pairs) to be miscounted during
line wrapping. Now uses Intl.Segmenter to properly handle multi-codepoint
characters.
Removed confusing parenthetical that suggested context files aren't loaded when continuing/resuming sessions. They are always freshly loaded from disk at startup - the parenthetical only applied to console logging behavior.
Working Indicator:
- Add '...' to channel messages while mom is processing
- Automatically removed when work completes or stops
- Applies to working message, not status messages
Improved Stop Command:
- Posts separate 'Stopping...' message that updates to 'Stopped'
- Original working message continues updating with tool results
- Clean separation between status and work output
- Properly handles abort during multi-step operations
Clean Stop Reason Handling:
- Agent run() now returns { stopReason } instead of throwing
- Handle 'aborted', 'error', 'stop', 'length', 'toolUse' cases cleanly
- No more exception-based control flow
- Track stopReason from assistant message.stopReason field
New SlackContext Methods:
- replaceMessage() - replace message text instead of appending
- setWorking() - add/remove working indicator
- Improved context tracking for stop command updates
Major improvements to mom's logging and cost reporting:
Centralized Logging System:
- Add src/log.ts with type-safe logging functions
- Colored console output (green=user, yellow=mom, dim=details)
- Consistent format: [HH:MM:SS] [context] message
- Replace scattered console.log/error calls throughout codebase
Usage Tracking & Cost Reporting:
- Track tokens (input, output, cache read/write) and costs per run
- Display summary at end of each run in console and Slack thread
- Example: 💰 Usage: 12,543 in + 847 out (5,234 cache read) = $0.0234
Prompt Caching Optimization:
- Move recent messages from system prompt to user message
- System prompt now mostly static (only changes with memory files)
- Enables effective use of Anthropic's prompt caching
- Significantly reduces costs on subsequent requests
Model & Cost Improvements:
- Switch from Claude Opus 4.5 to Sonnet 4.5 (~40% cost reduction)
- Fix Claude Opus 4.5 cache pricing in ai package (was 3x too expensive)
- Add manual override in generate-models.ts until upstream fix merges
- Submitted PR to models.dev: https://github.com/sst/models.dev/pull/439
UI/UX Improvements:
- Extract actual text from tool results instead of JSON wrapper
- Cleaner Slack thread formatting with duration and labels
- Tool args formatting shows paths with offset:limit notation
- Add chalk for colored terminal output
Dependencies:
- Add chalk package for terminal colors
- Add MEMORY.md files for persistent working memory
- Global memory: workspace/MEMORY.md (shared across channels)
- Channel memory: workspace/<channel>/MEMORY.md (channel-specific)
- Automatically loaded into system prompt on each request
- Enhance JSONL log format with ISO 8601 dates
- Add 'date' field for easy grepping (e.g., grep '"date":"2025-11-26"')
- Migrated existing logs to include date field
- Improve log query efficiency
- Add jq query patterns to prevent context overflow
- Emphasize limiting NUMBER of messages (10-50), not truncating text
- Show full message text and attachments in queries
- Handle null/empty attachments with (.attachments // [])
- Optimize system prompt
- Add current date/time for date-aware operations
- Format recent messages as TSV (43% token savings vs raw JSONL)
- Add efficient query examples with both JSON and TSV output
- Enhanced security documentation
- Add prompt injection risk warnings
- Document credential exfiltration scenarios
- Provide mitigation strategies
In -p, --mode json, and --mode rpc modes, don't print informational
messages like 'Loaded project context from:' or model restore messages.
Only the actual output should be printed.
BREAKING CHANGE: Passing a prompt on the command line now starts interactive
mode with the prompt pre-submitted, instead of exiting after completion.
Use --print or -p to get the previous non-interactive behavior.
- Add --print / -p flag for non-interactive mode
- Update runInteractiveMode to accept initial messages
- Update README documentation
- Fix Model Selection Priority docs to include --models scope
When pressing Enter on a highlighted slash command suggestion (e.g., typing
`/mod` with `/model` highlighted), the completion is now applied before
submitting. Previously, the partial text was submitted instead of the
selected command.
Fixes#49
PR #47 enhancements:
- Add thinking level syntax to --models (e.g., --models sonnet:high,haiku:low)
- First model in scope used as initial model when starting new session
- Auto-apply thinking level when cycling models with Ctrl+P
- Save both model and thinking to session AND settings for persistence
- Simplify UX by removing autoThinkingDisabled flag
- Fix model matching to prioritize exact matches over partial
- Support provider/modelId format (e.g., openrouter/openai/gpt-5.1-codex)
Issue #45:
- Add --thinking CLI flag to set thinking level directly
- Takes highest priority over all other thinking level sources
Closes#45
- Fixed catastrophic regex backtracking in extractPathPrefix that caused
terminal to hang when text contained many / characters (e.g., URLs).
Replaced complex regex with simple lastIndexOf approach. (#18)
- Fixed arrow keys moving both autocomplete selection and editor cursor
by adding missing return statement after handling arrow keys in
autocomplete mode.
Closes#18
- Add exact match support with '/' prefix (/gpt-5.1-codex or /provider/model)
- Prefer exact ID matches over partial matches in fuzzy search
- Parse thinking levels from --models flag (pattern:level format)
- Use first scoped model as initial model with its thinking level
- Auto-apply thinking when cycling with Ctrl+P
- Track manual thinking changes to disable auto-switching
- Clear model scope when using /model command
- Support mixed configs: --models sonnet:high,haiku,opus:low
- Silently ignore thinking for models that don't support it
- Only allow Ctrl+P cycling when --models is explicitly provided
- Update help text with examples
- Improve code organization by resolving scope early
- Consolidated theme system into single src/theme/ directory
- Created Theme class with fg(), bg(), bold(), italic(), underline()
- Added dark and light built-in themes with 36 color tokens
- Support for custom themes in ~/.pi/agent/themes/*.json
- JSON schema for theme validation
- Theme selector UI with /theme command
- Save theme preference to settings
- Uses chalk for text formatting to preserve colors
TODO:
- Replace hardcoded colors throughout TUI components
- Apply markdown theming to Markdown components
- Add theme support to all TUI elements
- Show offset/limit in read tool display (e.g., read src/main.ts:100-200)
- Fix PI_CODING_AGENT_DIR env var name in help and code
- Add all API key env vars to help text
The prepublish hook was running during npm install, before optional platform-specific dependencies were installed. This caused build failures on Linux for packages like @typescript/native-preview-linux-x64 and @parcel/watcher-linux-x64-glibc.
prepublishOnly only runs before npm publish, not during install.
Fixes#41
Fixes#39
- Added headers field to Model type (provider and model level)
- Model headers override provider headers when merged
- Supported in all APIs:
- Anthropic: defaultHeaders
- OpenAI (completions/responses): defaultHeaders
- Google: httpOptions.headers
- Enables bypassing Cloudflare bot detection for proxied endpoints
- Updated documentation with examples
Also fixed:
- Mistral/Chutes syntax error (iif -> if)
- process.env.ANTHROPIC_API_KEY bug (use delete instead of = undefined)
