From e4442faf55ceaf1fc68d3ac11cc1ca84d7bb0742 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E2=80=9CSebastian?= <64795732+slegarraga@users.noreply.github.com> Date: Thu, 6 Aug 2026 04:45:35 -0400 Subject: [PATCH] docs: document concurrency, junit and degrade flags --- README.md | 5 ++++- src/cli/index.ts | 3 +++ 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index b66ca8d..acc5ee8 100644 --- a/README.md +++ b/README.md @@ -204,9 +204,12 @@ evalgate init [file] Write a starter suite you can edit. evalgate list List the available scorers and providers. Flags: --provider --model --tags --concurrency --out --md --junit --json - --tolerance --comment --no-fail + --degrade --tolerance --comment --no-fail ``` +Try `evalgate run suite.eval.yaml`, then rerun with `--degrade` to watch the +exit code flip to 1 when the mock provider returns degraded output. + ## Library evalgate is also a library: diff --git a/src/cli/index.ts b/src/cli/index.ts index 7ec9b70..b854411 100644 --- a/src/cli/index.ts +++ b/src/cli/index.ts @@ -40,7 +40,10 @@ Common flags: --tags Only run cases with one of these tags. --out Write the JSON result artifact here. --md Write a Markdown report here. + --junit Write a JUnit XML report here. --json Print the JSON result to stdout. + --concurrency Run cases concurrently (default: 1). + --degrade Make the mock provider return degraded output. --no-fail Do not exit non-zero on failure/regression. compare flags: