# Review Public Boundary

> Reviews material before public exposure when asked to check for secrets, private identifiers, proprietary text, or unsupported claims; keeps private context out of public artifacts.

- Skill: `jovanipink/review-public-boundary` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx get-skill add jovanipink/review-public-boundary`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jovanipink/review-public-boundary/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: JovaniPink (https://skillmd.com/u/jovanipink)
- Updated: 2026-11-15
- Page: https://skillmd.com/skills/jovanipink/review-public-boundary

---


# Public-Private Boundary Review

Review both content and provenance. Renaming an internal identifier does not make proprietary logic, data, or operational knowledge public.

## Workflow

1. Define the intended public artifact, audience, license, and source repositories.
2. Inventory copied, adapted, generated, and independently authored material.
3. Check license and authorization for every non-original source. Treat absent permission as a blocker, not an invitation to sanitize.
4. Scan content and history in scope for:
   - credentials, tokens, keys, cookies, connection strings, and personal data
   - private repository URLs, absolute local paths, account IDs, project IDs, hostnames, emails, and internal ticket references
   - client names, proprietary schemas, business rules, architecture contracts, incidents, logs, screenshots, and production configuration
   - claims that confuse repository code with testing, deployment, availability, or outcomes
5. Recommend a disposition. Replace only ordinary incidental identifiers. For proprietary substance, remove it and independently author from public sources or keep it in a private overlay. Edit files only when the request asks for redaction.
6. After an authorized redaction, re-run tests and documentation checks so safety edits do not leave broken artifacts.
7. Report blockers, safe-to-publish items, and facts requiring owner or counsel confirmation.

## Output

Lead with the answer. Use the shortest format that retains the required evidence and limits; expand when requested or needed. The items below describe required content, not mandatory headings.

Use `BLOCK`, `REVIEW`, or `CLEAR` per finding, with location, risk, evidence, and disposition. A `CLEAR` verdict means no issue was found in the reviewed scope; it is not a legal opinion or proof that unreviewed history is safe. Never quote a secret in a finding: give its type and location, and mask the value.

## Example

Synthetic example; not an observed client result.

Input: A draft guide includes an internal dataset identifier.

Output: Remove or generalize the private identifier before public exposure. Keep any needed operational detail in its owning private context.

