Feature/ai nl parser - #15
Merged
Merged
Conversation
Why: - Reduce friction of remembering taskwarrior attribute syntax by letting users describe tasks in plain language What: - Add tools/ai.py with provider-agnostic parse_nl_to_command() function - Support Anthropic (claude-sonnet-4-6) and OpenAI (gpt-4o-mini) backends - Late-import both SDKs so they remain optional dependencies - Add anthropic and openai as optional deps under [extras] ai in pyproject.toml Impact: - No breaking change; providers only activate when explicitly used Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why: - Wire the AI parser into the CLI so users can run `twf ai "buy groceries tomorrow high priority"` directly What: - Add ai() command to main.py accepting free-form prompt as positional args - --group / -g selects taskwarrior group (defaults to first in config) - --provider / -p overrides AI provider (falls back to TW_AI_PROVIDER env, then "anthropic") - Reuses existing confirmation + subprocess execution pattern from task_wrapper - Import group_mappings_completion for --group autocompletion Impact: - No breaking change to existing commands Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why:
- Users shouldn't have to set env vars just to use AI; TW_CONFIG is the
existing single source of truth for all TWF settings
What:
- Add 'ai' section to default TW_CONFIG (provider, anthropic_api_key, openai_api_key)
- Update _parse_with_* to accept config_api_key; env var still wins over config
- Resolution order: CLI flag > env var > TW_CONFIG > default ("anthropic")
- Pass config_api_key through parse_nl_to_command to ai command in main.py
- Update error messages to reference both env var and TW_CONFIG options
Impact:
- No breaking change; existing env var usage continues to work unchanged
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why: - UV is significantly faster for installs and lock resolution, and hatchling is the standard PEP 517 backend for uv-managed projects What: - Replace [tool.poetry] with [project] (PEP 621); hatchling build backend - Move scripts/extras/dev-deps to [project.scripts], [project.optional-dependencies], [dependency-groups] respectively - Add [tool.hatch.build.targets.wheel] packages = ["tools"] for package discovery - Remove poetry.lock (run `uv sync` to generate uv.lock) - Update ci.yml: replace snok/install-poetry with astral-sh/setup-uv, use uv run for ruff/pytest, bump test matrix to 3.10/3.11/3.12 - Update python-publish.yml: use uv build + uv publish with UV_PUBLISH_TOKEN Impact: - Run `uv sync` to generate uv.lock; `uv sync --extra ai` for AI providers - pipx install still works unchanged via PEP 517 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why: - Old README had minimal installation info and no mention of AI features or the UV migration What: - Add install instructions for standard and [ai] extra via pipx and uv - Add configuration section documenting TW_CONFIG structure and AI key resolution order (CLI flag > env var > TW_CONFIG > default) - Add AI natural language section with example inputs/outputs and option table - Add development section with uv sync commands - Fix related tools section Impact: - Documentation only; no code change Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why: - Complete the UV migration: uv.lock is the new source of truth for reproducible installs Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why: - Version reflects the new AI feature and UV migration - Config default path had a typo (confg.json → config.json) - README needed the updated default path and mtwd integration instructions What: - pyproject.toml: 0.5.2 → 0.6.0 - tools/__init__.py: fix typo confg.json → config.json - README: update default config path to ~/.local/share/tw_flow/config.json, simplify default config example, add TW_CONFIG export instructions for m_taskwarrior_d.nvim users (bash/zsh/fish) Impact: - Existing users with TW_CONFIG set are unaffected - New users get config at ~/.local/share/tw_flow/config.json Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why: - No external issue tracker is used for this project, so requiring issue links on TODOs adds noise with no benefit Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Why: - Simplifying the default config broke tests that assumed the old 'task' group and 'Bills' template existed; also exposed a typo (confg.json) and a chicken-and-egg import ordering bug in the test setup What: - Add root conftest.py: pytest_configure hook writes a fixed TEST_CONFIG (personal + test groups, Bills template, Test project query) to TW_CONFIG before any tools module is imported - tests/conftest.py: move tools imports into fixtures so they happen after pytest_configure has written the config file - pyproject.toml: ignore ruff TD003 (missing issue link) — no issue tracker - tools/__init__.py: fix typo confg.json → config.json in default path Impact: - Tests now pass on a fresh environment with no pre-existing config file Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.