From 06578d941dc9eae5c5514dcda729e03a67bd026c Mon Sep 17 00:00:00 2001 From: James Russo Date: Tue, 17 Nov 2026 22:12:00 +0000 Subject: [PATCH] chore(github): secret scanning skips the generated skills manifest (#8115) Its per-skill content hashes are read as passwords, and each regeneration raises a new alert. Co-authored-by: Claude Opus 5.5 --- .github/secret_scanning.yml | 7 +++++++ 1 file changed, 7 insertions(+) create mode 100644 .github/secret_scanning.yml diff --git a/.github/secret_scanning.yml b/.github/secret_scanning.yml new file mode 100644 index 0000000000..a4f08822b7 --- /dev/null +++ b/.github/secret_scanning.yml @@ -0,0 +1,7 @@ +# Generic (AI-detected) secret scanning reads the generated skills manifest's per-skill +# 16-hex content hashes as passwords, and every regeneration raises new alerts. The file +# is written by packages/cli/scripts/gen-skills-manifest.ts and holds only skill names, +# content hashes and file counts. Provider-pattern scanning and push protection still +# cover every other path. +paths-ignore: + - "skills-manifest.json"