Breaking change: Pi skills must now be named SKILL.md inside a directory,
matching Codex CLI format. Previously any *.md file was treated as a skill.
Migrate by renaming ~/.pi/agent/skills/foo.md to ~/.pi/agent/skills/foo/SKILL.md
* Add skills system with Claude Code compatibility
* consolidate skills into single module, merge loaders, add <available_skills> XML tags
* add Codex CLI skills compatibility, skip hidden/symlinks
- Add Mistral to KnownProvider type and model generation
- Implement Mistral-specific compat handling in openai-completions:
- requiresToolResultName: tool results need name field
- requiresAssistantAfterToolResult: synthetic assistant message between tool/user
- requiresThinkingAsText: thinking blocks as <thinking> text
- requiresMistralToolIds: tool IDs must be exactly 9 alphanumeric chars
- Add MISTRAL_API_KEY environment variable support
- Add Mistral tests across all test files
- Update documentation (README, CHANGELOG) for both ai and coding-agent packages
- Remove client IDs from gemini.md, reference upstream source instead
Closes#165
- Add agentLoopContinue() to pi-ai for resuming from existing context
- Add Agent.continue() method and transport.continue() interface
- Simplify AgentSession compaction to two cases: overflow (auto-retry) and threshold (no retry)
- Remove proactive mid-turn compaction abort
- Merge turn prefix summary into main summary
- Add isCompacting property to AgentSession and RPC state
- Block input during compaction in interactive mode
- Show compaction count on session resume
- Rename RPC.md to rpc.md for consistency
Related to #128
Sanitize shell output by removing Unicode Format characters and lone
surrogates that crash string-width. This fixes crashes when running
commands like curl that download binary files.
- Added totalTokens field to Usage interface in pi-ai
- Anthropic: computed as input + output + cacheRead + cacheWrite
- OpenAI/Google: uses native total_tokens/totalTokenCount
- Fixed openai-completions to compute totalTokens when reasoning tokens present
- Updated calculateContextTokens() to use totalTokens field
- Added comprehensive test covering 13 providers
fixes#130
- Updated isBunBinary detection to check for %7EBUN (URL-encoded ~BUN)
- Simplified build:binary script to remove Unix-specific shell syntax
- Binary now correctly locates supporting files next to executable on Windows
Since version 0.9.0, RPC mode (--mode rpc) was not saving messages to
session files. The agent.subscribe() call with session management logic
was only present in the TUI renderer after it was refactored.
RPC mode now properly saves sessions just like interactive mode.
Added test for RPC mode session management to prevent regression.
Fixes#83
Thanks @kiliman for reporting this issue!
Breaking Changes:
- Timestamps now use Slack format - run migrate-timestamps.ts for existing logs
Added:
- Channel/user ID mappings in system prompt
- Skills documentation in system prompt
- Debug last_prompt.txt output
- Bash working directory info
- Token-efficient log queries filtering tool calls
Changed:
- Turn-based message context (groups consecutive bot messages as one turn)
- Messages sorted by Slack timestamp
- Condensed system prompt (~5k → ~2.7k chars)
- Simplified user prompt
- Selective logging (skip UI status labels)
Fixed:
- Duplicate message logging from app_mention handler
- Username obfuscation in thread messages
- Fix line numbers showing incorrect values for edits far from file start
(e.g., 1,2,3 instead of 336,337,338). Skip count was added after displaying
lines instead of before.
- Rewrite splitIntoTokensWithAnsi in pi-tui to preserve whitespace as separate
tokens instead of discarding it. Wrapped lines now maintain proper alignment
and code indentation.
- Update mom README: rename title, remove em-dashes for cleaner prose
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
- 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
Mario Zechner