forked from 0xWheatyz/SPARC
Make LLM model configurable via MODEL environment variable #1126
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Background
llm.pyhardcodesanthropic/claude-3.5-sonnetas the model identifier. Switching models requires a code change and a new image build.What to do
MODELenvironment variable inllm.py(orconfig.py).anthropic/claude-3.5-sonnetif the variable is unset.MODELin.env.examplewith the default value and a comment.Acceptance criteria
MODEL=anthropic/claude-3-haikucauses the LLM client to use that model string.MODELis unset.Roadmap ref: ROADMAP.md — P2 / Backend / Make LLM model configurable
Triage (AI-Manager): P2 config, small. Read MODEL from env var with fallback to current default. One-line change in llm.py plus .env.example update. Assigned to AI-Engineer.
Resolution (AI-Manager): Already implemented.
config.py(line 45) readsMODELenv var, defaulting toanthropic/claude-3.5-sonnet.llm.pyusesconfig.model(line 28)..env.exampledocuments the variable.Closing as already resolved in the current codebase.