feat: deepseek-v4-flash default + terminal REPL#23
Merged
Conversation
Set llm_provider=deepseek and model=deepseek-v4-flash as the built-in defaults (config.py) and in config.yml.example.
Adds core/repl.py and a 'make repl' target. Builds the agent from the shared config store, registers a 'repl' channel that routes permission approvals to a y/n stdin prompt, and loops on stdin — lets you exercise the agent end to end without Telegram.
Capture model reasoning (anthropic thinking blocks, deepseek reasoning_content) on LLMResponse and emit it via a dedicated core.llm.reasoning logger, silent by default. The REPL bumps that logger plus core.agent/core.executor to INFO and streams them live above a \r spinner, so you see the reasoning and every tool call as the agent works.
- Guard redraw() so startup/idle log records can't draw the spinner with a bogus elapsed time (the '…3128121s' line). - Print a debug config block at launch: active provider/model, thinking level, memory model, history mode, voice, timezone. - Tool-call/agent activity now renders cyan, reasoning stays dim, so the chain of thought reads at a glance.
- ESC during a turn cancels the in-flight agent.process task (POSIX tty via cbreak + add_reader; lone ESC only, escape sequences ignored). - /clear wipes this session's conversation history. - Banner documents both.
mattmezza
force-pushed
the
feat/repl-deepseek-default
branch
from
8749f29 to
ece518a
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
deepseek-v4-flash(providerdeepseek) inconfig.pydefaults andconfig.yml.example.core/repl.py,make repl): chat with the agent locally without Telegram. Builds the agent from the shared config store, registers areplchannel that routes permission approvals to a y/n stdin prompt, and loops on stdin.Test
make repl→ type a message. Ctrl-D or/exitquits.316 tests pass.