菜单
Engineer privacy: build executable data-subject-rights pipelines (access/erasure/portability) that locate data via the catalog, track lawful basis + granular revocable consent, minimize collection, automate retention/deletion, and distinguish pseudonymization from anonymization.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/mcorbett51090/RavenClaude --skill privacy-mechanics命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Route everyday, well-defined work (single-file edits, test-writing, summarization, mechanical refactors, boilerplate) to a cheaper coding agent (Grok or Copilot CLI) instead of doing it in the main Claude session, escalating multi-file/architecture/debugging/security work to Claude. Opt-in via the `cheap_lane:` comfort-posture knob.
Authors, validates and packages a claude.ai Organization Skill — the intake questions, the trigger and scope exercises, then lint, pack and verify against measured platform constraints. Reach for this when someone wants to publish a Skill to their organization, or when an upload was rejected and the reason is unclear.
Drafts a reply to an inbound customer email, grounded in the account record and the published help-centre articles rather than in recall. Use when a customer emails about billing, a renewal or an outage, when a reply must cite current policy, or when a teammate asks for a first draft to edit.
基于 SOC 职业分类
正在显示 SKILL.md
| name | privacy-mechanics |
| description | Engineer privacy: build executable data-subject-rights pipelines (access/erasure/portability) that locate data via the catalog, track lawful basis + granular revocable consent, minimize collection, automate retention/deletion, and distinguish pseudonymization from anonymization. |
Access/erasure/portability executed across every system holding the person's data (located via catalog/lineage). 'Couldn't find it all' = failed DSR.
Each use has a recorded basis; consent (where it's the basis) is granular, timestamped, revocable, and revocation propagates.
Challenge every field (need it? how long? what basis?). Automate retention/deletion — indefinite retention is unbounded risk.
Pseudonymized (re-identifiable with a key) is still personal data. True anonymization is a high bar. Name which you achieved. Legal interpretation -> route out.
Runnable engine (reversible). To actually pseudonymize text before it reaches a model — reversibly, restored on the way back — use ravenclaude-core's pseudonymize skill (scripts/pseudonymize.py): a deterministic known-entity denylist (the reliable core) + structured-PII regex + optional NER, with a local key vault. It enforces exactly the caveat above — the vault is the re-identification key, so the output is still personal data; keep the vault local (0600, out of git) and never send it to the model. It states its honest ceiling plainly: it reliably removes the entities you list, not every free-text name, and never claims anonymity (any unlisted variant leaks; quasi-identifiers re-identify even with names gone).