Release v0.1.6

Highlights

• ScriptedTool for composing multi-tool bash orchestration with Python/LangChain bindings
• Streaming output support for Tool trait
• Script file execution by path
• 10 interpreter bug fixes surfaced by eval harness

What's Changed

• chore: pre-release maintenance checklist (#223) by @chaliy
• feat(interpreter): support executing script files by path (#222) by @chaliy
• fix(jq): fix argument parsing, add test coverage, update docs (#221) by @chaliy
• feat(tool): add streaming output support (#220) by @chaliy
• feat(python): ScriptedTool bindings + LangChain integration (#219) by @chaliy
• refactor(examples): extract fake tools into separate module (#218) by @chaliy
• chore: add small-PR preference to AGENTS.md (#217) by @chaliy
• fix(builtins): resolve 10 eval-surfaced interpreter bugs (#216) by @chaliy
• fix: address 10 code TODOs across codebase (#215) by @chaliy
• test: add skipped tests for eval-surfaced interpreter bugs (#214) by @chaliy
• feat(scripted_tool): add ScriptedTool for multi-tool bash composition (#213) by @chaliy
• ci(python): add Python bindings CI with ruff and pytest (#212) by @chaliy
• fix(interpreter): apply brace/glob expansion in for-loop word list (#211) by @chaliy
• feat(python): add PydanticAI integration and example (#210) by @chaliy
• fix(ci): add --allow-dirty for cargo publish after stripping monty (#209) by @chaliy
• fix(ci): strip git-only monty dep before crates.io publish (#208) by @chaliy

Full Changelog: v0.1.5...v0.1.6