gongzhiyongandClaude Opus 4.8
1288fd19d7
按 juejin 协作聚合模型重构收敛(取代 best-of-N 选优): - 删蜂后选优(queen.py/test-queen.py) - 新增聚合节点 result_aggregator.py:共享池收集→同文件 LLM/AST 整合→沙箱验证→单次落 main - 质量驱动闭环:不达标打回迭代(AGGREGATE_ACCEPTANCE_THRESHOLD + MAX_REVIEW_CYCLES) - agent 停 git 工作分支,产出走 task.result.files 共享池(AGENT_GIT_PUSH_ENABLED 默认 false) - sandbox_runner 改用 pytest(原生支持 pytest 风格 class),修 stdlib runner 收集失败 - 文档同步重写为协作聚合模型 本地验证:产物仓单分支 main + 三函数完整 + pytest 12/12 pass_rate=100 一次达标。 影响:Swarm 收敛/聚合层;Manager/客户端契约不变(artifact字段/sequence/状态机;契约测试全过)。 Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
28 lines
1.1 KiB
Plaintext
28 lines
1.1 KiB
Plaintext
fastapi==0.115.0
|
|
uvicorn[standard]==0.32.0
|
|
websockets==13.1
|
|
redis==5.2.0
|
|
openai==1.55.3
|
|
pydantic==2.9.2
|
|
python-dotenv==1.0.1
|
|
prometheus-client==0.20.0
|
|
httpx==0.28.1
|
|
# dev/CI only: in-memory Redis emulator for the gated REDIS_FAKE/ALLOW_MEMORY_STORE fallback
|
|
fakeredis==2.26.1
|
|
# sandbox test runner: pytest natively runs the test styles agents produce (pytest-style classes,
|
|
# fixtures, parametrize, marks) — the in-pod harness (sandbox_runner.py) prefers it, stdlib fallback.
|
|
pytest==8.3.3
|
|
opentelemetry-api==1.24.0
|
|
opentelemetry-sdk==1.24.0
|
|
opentelemetry-exporter-otlp==1.24.0
|
|
opentelemetry-instrumentation-redis==0.45b0
|
|
opentelemetry-instrumentation-requests==0.45b0
|
|
opentelemetry-instrumentation-logging==0.45b0
|
|
# benchmark metrics export targets (lazy-imported; only used when BENCHMARK_EXPORT_TARGET set)
|
|
azure-cosmos==4.7.0
|
|
azure-storage-blob==12.23.1
|
|
azure-identity==1.19.0
|
|
# model-key resolution from Key Vault via Pod workload identity (lazy-imported in
|
|
# agent_launcher._resolve_from_keyvault; only when AZURE_FEDERATED_TOKEN_FILE / SECRET_RESOLVER=azkv)
|
|
azure-keyvault-secrets==4.9.0
|