Skip to content

Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.

Chore/pr agent ark token billing (#1117)

bot-of-qin-ctxe8c129c51fba
diff --git a/.github/workflows/pr-review.yml b/.github/workflows/pr-review.yml
index 7ffb812e7..8f8570b88 100644
--- a/.github/workflows/pr-review.yml
+++ b/.github/workflows/pr-review.yml
@@ -25,8 +25,7 @@ jobs:
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
OPENAI_KEY: ${{ secrets.DOUBAO_API_KEY }}
- OPENAI.API_BASE: "https://ark.cn-beijing.volces.com/api/coding/v3"
- config.model: "ark-code-latest"
+ OPENAI.API_BASE: "https://ark.cn-beijing.volces.com/api/v3"
github_action_config.auto_review: "true"
- github_action_config.auto_describe: "true"
+ github_action_config.auto_describe: "false"
github_action_config.auto_improve: "true"
diff --git a/.pr_agent.toml b/.pr_agent.toml
index 2c7ba7310..5396cecba 100644
--- a/.pr_agent.toml
+++ b/.pr_agent.toml
@@ -17,6 +17,10 @@
# ---------------------------------------------------------------------------
[config]
output_relevant_configurations = false
+# Use Ark token-billed Chat API instead of Coding Plan.
+model = "doubao-seed-2-0-code-preview-260215"
+fallback_models = []
+custom_model_max_tokens = 256000
# Use high reasoning for this complex polyglot codebase with subtle async/
# concurrency bugs. Worth the extra latency for quality.
reasoning_effort = "high"
diff --git a/tests/agfs/conftest.py b/tests/agfs/conftest.py
index 0b41a614e..20dbe1f35 100644
--- a/tests/agfs/conftest.py
+++ b/tests/agfs/conftest.py
@@ -6,7 +6,6 @@ from pathlib import Path
import pytest
-
@pytest.fixture(scope="session")
def agfs_test_root():
"""Root directory for AGFS tests."""