Conversation
…_batch Replace rich-styled output with plain print/stdout.write, eliminating the display.py module and the rich dependency entirely. Inline the ChatClient from chat.py. Remove unused execute_batch() and its test. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Rich is only used for Live/Spinner during tool execution — all other output stays as plain print/stdout.write. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Merge main.py and models.py into orchestrator.py and tools.py, replacing Pydantic models with plain dicts and tenacity retry with a simple loop. Removes 3 dependencies (pydantic, pydantic-settings, tenacity). Updates tests and docs to match the new interface. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Keep slim-display's 2-module layout (no pydantic/tenacity/display). Pull in from main: generic LLM_API_KEY/LLM_BASE_URL config and the research empty-response fix. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Roll back history to a pre-turn snapshot when _stream_response returns None, so an LLM/API failure after the user message (or after partial tool-call/result pairs) doesn't leave dangling entries that would poison the next turn. The Ctrl+C-with-partial-content path is unchanged — that reflects user intent, not infra failure. Also update DISCOVERIES.md to match the post-pydantic/tenacity code: _format_weather shape detection, plain-dict tool results, _race_with_cancel / _wait_or_cancel, and RuntimeError for exhausted throttle retries. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Pull DISCOVERIES.md changes from main (removes the Application-Level Issues section). Keep slim-display's architecture untouched: flat async functions in orchestrator.py, no models.py, no display/main modules. The stale-reference fixes already on slim-display are preserved because they describe code that only exists on this branch. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
# Conflicts: # src/cli_chat/orchestrator.py
Inline the two history.append loops in _process_turn and default _tool_result's error flag to False so success sites drop the positional bool. Net fewer lines and clearer call sites. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two cherry-picks from main, adapted to slim-display's architecture: - Bump REQUEST_TIMEOUT from 15s to 20s so the research API's 15s slow path always finishes server-side (DISCOVERIES.md), instead of racing our httpx.ReadTimeout — which can wrap an empty asyncio.TimeoutError and surface as a cryptic "Request failed: " error. - Wrap the catch-all httpx exception handler to include the exception class name when str(exc) is empty, so future empty-message exceptions remain diagnosable. - Reword README "Requirements" to list env vars as the primary interface, noting that .env is one option (picked up automatically by uv run). Old wording implied keys had to live in .env. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
CLAUDE.md still framed .env as the required key store; README now treats env vars as the primary interface with .env as one option (picked up by uv run). Update CLAUDE.md to match. Co-Authored-By: Claude Opus 4.6 (1M context) <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.
Summary
display.py— inlined all display logic intoorchestrator.pyas plainprint/sys.stdout.writechat.pyintoorchestrator.py(same as Merge chat.py into orchestrator, genericize system prompt #1, also included here for independence)richdependency (and its transitive depsmarkdown-it-py,mdurl)execute_batch()fromToolExecutorand its testTest plan
make checkpasses — ruff, pyright, pylint all clean, 20/20 tests passmake runand verify streaming + tool calling still works🤖 Generated with Claude Code