graff — a fast agentic coding harness in Zig: multi-provider, MCP, workflows, DGM evolution loop, TS/Python SDKs
Python
Gemini direct: apply the route-wide effort policy
justrach3f0270d52804
diff --git a/src/interactions_wire.zig b/src/interactions_wire.zigindex 8bf5d96f..f72a28f1 100644--- a/src/interactions_wire.zig+++ b/src/interactions_wire.zig@@ -60,7 +60,9 @@ pub fn write(self: *Agent, s: *std.json.Stringify, tools: ?[]const u8, force_tootry s.objectField("generation_config");try s.beginObject();try s.objectField("thinking_level");- try s.write(thinkingLevel(@tagName(self.reasoning)));+ // Same effort policy as the Codegraff route (ADR 0046: default medium on+ // Gemini goes out as low), so both paths think alike.+ try s.write(thinkingLevel(@import("effort_route.zig").wireEffort(self.provider.id, self.provider.model, @tagName(self.reasoning))));// Documented as the default, but summaries only stream when asked for;// they feed the reasoning panel (title.reasoningDelta).try s.objectField("thinking_summaries");