From 582dbbbbf7c4cc241dbb7bcbdd7cd80e6751a798 Mon Sep 17 00:00:00 2001 From: Teknium <127238744+teknium1@users.noreply.github.com> Date: Mon, 6 Apr 2026 11:22:07 -0700 Subject: [PATCH] feat: add grok to TOOL_USE_ENFORCEMENT_MODELS for direct xAI usage (#5595) Grok models (x-ai/grok-4.20-beta, grok-code-fast-1) now receive tool-use enforcement guidance, steering them to actually call tools instead of describing intended actions. Matches both OpenRouter (x-ai/grok-*) and direct xAI API usage. --- agent/prompt_builder.py | 2 +- tests/agent/test_prompt_builder.py | 3 +++ 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/agent/prompt_builder.py b/agent/prompt_builder.py index 80af3b64d..0a2cbe374 100644 --- a/agent/prompt_builder.py +++ b/agent/prompt_builder.py @@ -187,7 +187,7 @@ TOOL_USE_ENFORCEMENT_GUIDANCE = ( # Model name substrings that trigger tool-use enforcement guidance. # Add new patterns here when a model family needs explicit steering. -TOOL_USE_ENFORCEMENT_MODELS = ("gpt", "codex", "gemini", "gemma") +TOOL_USE_ENFORCEMENT_MODELS = ("gpt", "codex", "gemini", "gemma", "grok") # OpenAI GPT/Codex-specific execution guidance. Addresses known failure modes # where GPT models abandon work on partial results, skip prerequisite lookups, diff --git a/tests/agent/test_prompt_builder.py b/tests/agent/test_prompt_builder.py index ce8084709..17e3523c0 100644 --- a/tests/agent/test_prompt_builder.py +++ b/tests/agent/test_prompt_builder.py @@ -1018,6 +1018,9 @@ class TestToolUseEnforcementGuidance: def test_enforcement_models_includes_codex(self): assert "codex" in TOOL_USE_ENFORCEMENT_MODELS + def test_enforcement_models_includes_grok(self): + assert "grok" in TOOL_USE_ENFORCEMENT_MODELS + def test_enforcement_models_is_tuple(self): assert isinstance(TOOL_USE_ENFORCEMENT_MODELS, tuple)