Bundle
dsh-eval-harness
DSH 插件回归评测门禁:yaml 用例 + headless 驱动 + trace 断言 + baseline 门禁(eval_run / eval_gate)
BiBoyang13
6 results
DSH 插件回归评测门禁:yaml 用例 + headless 驱动 + trace 断言 + baseline 门禁(eval_run / eval_gate)
Deterministic, CI-safe golden-output evaluation for DeepSeek Harness
Configurable user-confirmed lessons that help DeepSeek Harness avoid repeating mistakes.
Proxy-egress WebFetchProvider for ctx.web: routes web_fetch through one explicit HTTP(S) proxy (undici ProxyAgent) instead of local DNS pinning, for fake-ip/TUN DNS environments. No core patches.
Run one command N rounds and judge by median/distribution instead of a single run
DeepSeek Harness plugin: a fail-closed URL host and scheme allowlist checked before a request is opened, exposed as assertUrlAllowed and the ssrf_check tool