safety-secret-detector

name	safety/secret-detector
description	Scan any export, file, or vault batch for credentials, API keys, tokens, and sensitive data before they leave the local machine. Fires automatically on all Claw export operations. Default skill for Starlight Sentinel and Attestation Claw.
type	safety

Secret Detector

Nothing sensitive exits the local machine without explicit human awareness.

When This Skill Activates

• Any content is about to be written outside ~/.starlight/ to an external destination
• Any vault export is generated for a platform adapter
• Any content is passed to an external MCP server
• Keywords: "export", "publish", "send", "upload", "share", "transmit"
• Default for: Starlight Sentinel

What This Skill Does

Scans content for credential patterns, personal identifiers, and sensitive data before it leaves the local machine. Blocks export if secrets are found. Never auto-redacts — always surfaces findings to the user for an explicit decision.

Procedures

Procedure 1: Pre-Export Scan

1. Receive content blob pending export

2. Run pattern detection against the full content:

   Pattern	Example	Action on match
   API keys	sk-..., pk_..., ghp_..., xoxb-...	BLOCK + report
   Private keys	-----BEGIN PRIVATE KEY-----	BLOCK + report
   Passwords in config	password=, passwd=, secret=	BLOCK + report
   Email addresses	[e-mail] in unexpected context	WARN + report
   Phone numbers	+1-555-..., formatted phone patterns	WARN + report
   Credit card patterns	16-digit sequences with separators	BLOCK + report
   Government IDs	SSN patterns, national ID formats	BLOCK + report

3. If BLOCK patterns found: halt export, generate SECRET_SCAN_REPORT.md, present to user

4. If WARN patterns found: present to user with option to proceed or redact

5. If clean: return APPROVED for export

Procedure 2: Scan Report Generation

1. Create /security/SECRET_SCAN_REPORT.md
2. Include: content source, patterns detected, line references (not the secrets themselves)
3. Present options: [redact and retry] / [cancel export] / [I confirm this is safe, proceed]
4. Log the decision with user action

Procedure 3: False Positive Handling

1. User declares a detected pattern is not sensitive
2. Record the declaration with timestamp + user confirmation
3. Add to local allowlist for this session (not persisted globally)
4. Proceed with export

Integration Points

• Vaults: No vault access needed — operates purely on content in transit
• Agents: Starlight Sentinel
• Skills: Composes with private-public-split for structural classification; permission-gate for the approval flow
• Claws: Called by every Claw that generates exports (Memory, Creator, Attestation, Reclamation)

Quality Criteria

• Zero false negatives on common credential patterns
• False positives surface clearly as warnings, not blocks (where safe)
• User is always in control — no silent redaction or silent pass
• Scan report is human-readable, not a raw regex dump