CloakClawv0.1.3

A law firm uses CloakClaw to redact party names, case numbers, and financial terms from contracts before sending them to a cloud LLM for clause analysis. After the model returns its review, CloakClaw restores the original entities so the attorney sees a fully contextualized response. This allows lawyers to leverage AI speed while maintaining client confidentiality.

An investment analyst pastes quarterly earnings transcripts containing company names, revenue figures, and account numbers into a cloud LLM. CloakClaw automatically cloaks all financial entities and proper nouns before the text leaves the local machine. The decloaked summary gives the analyst actionable insights without leaking proprietary data to third-party servers.

An HR department uses CloakClaw to redact candidate names, emails, phone numbers, and addresses from resumes before asking a cloud LLM to rank candidates by skills. The cloak-then-decloak workflow ensures the AI never sees real personal identifiers, reducing bias and privacy risk. The final ranked list restores original candidate details for the hiring manager.

A clinic uses CloakClaw's medical profile to strip patient names, medical IDs, and dates from clinical notes before sending them to an LLM for preliminary diagnosis summarization. The local regex and Ollama-based detection keeps PHI off cloud servers. Restored output gives doctors a readable summary while adhering to best-effort privacy practices.

A DevOps engineer uses CloakClaw's code profile to cloak API keys, passwords, IP addresses, and URLs in .env files and infrastructure configs before pasting them into a cloud LLM for troubleshooting. The AI suggests fixes without ever seeing real credentials. Decloaked responses allow the engineer to apply corrections directly to the actual configuration.

CloakClaw is distributed as a free, open-source CLI tool with core redaction features. A paid Enterprise tier adds team management, audit logs, SOC 2 compliance reporting, and priority support. This attracts individual developers while monetizing larger organizations with stricter compliance needs.

A hosted version of CloakClaw runs as a transparent proxy between the user and cloud LLM APIs. The service handles cloaking, decloaking, and session management without local installation. It targets teams that want privacy without managing infrastructure.

CloakClaw's engine is licensed to LLM platform providers, legal tech companies, and healthcare SaaS vendors to embed privacy redaction directly into their products. This allows partners to offer compliant AI features without building redaction from scratch.