fix(server): multi step tool call (#13486)

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- Bug Fixes
- Enforced a consistent step limit for AI responses across providers,
preventing excessively long generations in both text and streaming modes
for more predictable results.

- Refactor
- Centralized step limit configuration into a shared provider, ensuring
uniform behavior across providers and simplifying future maintenance.
- Standardized application of step limits in text generation and
streaming flows to align provider behavior and improve overall
reliability.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
This commit is contained in:
DarkSky
2025-08-13 16:10:42 +08:00
committed by GitHub
parent 7e0de251cb
commit dd9d8adbf8
4 changed files with 4 additions and 4 deletions
@@ -296,8 +296,6 @@ export class OpenAIProvider extends CopilotProvider<OpenAIConfig> {
},
];
private readonly MAX_STEPS = 20;
#instance!: VercelOpenAIProvider;
override configured(): boolean {