A Claude Code skill bundle for bug hunting and external red-team work - 82 skills, 15 slash commands, 681 disclosed-report patterns curated across 24 core vulnerability classes, plus enterprise identity + infrastructure attack matrices.
Python ai-securityanthropicapplication-securitybug-bountybugbountybugcrowdclaudeclaude-codeclaude-skillsethical-hackinghackeroneoffensive-securitypentestingred-teamsecurity-toolsweb-security
feat(grounding): enforce sources/report_count (GROUNDING_STRICT = True) (#69)
Sachin Sharmab7b2d8edef53
diff --git a/scripts/lint_skills.py b/scripts/lint_skills.pyindex cad1402..c82e64b 100644--- a/scripts/lint_skills.py+++ b/scripts/lint_skills.py@@ -42,8 +42,8 @@ MAX_BODY_LINES = 500# Grounding enforcement. Every skill must declare `sources:`; hunt-* skills (built# from disclosed reports) must also declare an integer `report_count:` (0 is fine# when `sources:` justifies it — e.g. a pattern library or research-derived skill).-# Phase 4a ships this as WARNINGS while the backfill lands; 4b flips it to errors.-GROUNDING_STRICT = False+# Phase 4a shipped this as WARNINGS while the backfill landed; now enforced as errors.+GROUNDING_STRICT = True# --- real-secret patterns (kept tight to avoid flagging documented regexes) ---SECRET_PATTERNS = [