Community寫作與編輯github.com

vizuh/clicktrail-claude-plugin

Claude Code plugin for consent-safe ClickTrail setup, instrumentation, audit, and verification

clicktrail-claude-plugin 是什麼?

clicktrail-claude-plugin is a Claude Code agent skill that claude Code plugin for consent-safe ClickTrail setup, instrumentation, audit, and verification.

相容平台✓Claude Code~Codex CLI~Cursor
npx skills add vizuh/clicktrail-claude-plugin

Installed? Explore more 寫作與編輯 skills: steipete/notion, langchain-ai/langchain, bytedance/podcast-generation · View all 6 →

在你喜歡的 AI 中提問

開啟一個已預先載入此 Agent Skill 的新對話。

說明文件

Audit ClickTrail

Remain read-only. Do not install packages, edit files, or call live endpoints.

  1. Read project instructions, tracking docs, manifests, and the integration entry points.
  2. Read ../../references/clicktrail-safety-contract.md.
  3. Trace startup, capture, persistence, delivery, withdrawal, stop, and retry paths. Inspect every caller of shared constructors or serializers.
  4. Check that trusted host configuration wins over caller-controlled attribution.
  5. Check server destinations for public credential-free HTTPS enforcement.
  6. Check payloads against a concrete allowlist. Treat arbitrary objects and raw passthrough as findings even when current callers appear safe.
  7. Check tests for observable consent-denial, withdrawal-erasure, redaction, SSRF, delivery-failure, and tenant-collision behavior.

Report findings as Critical, Important, or Minor. Each finding must name the file, behavior, impact, and smallest safe correction. End with verified passes and unresolved evidence. Do not turn missing evidence into a pass.

相關技能