vendor-ai-review

To review vendor AI terms against your governance policy, the skill performs term-by-term extraction comparing training, liability, model changes, and incident notification obligations against your configured CLAUDE.md playbook, producing actionable redlines.

A vendor AI agreement gap analysis detects conflicts between vendor terms and internal governance positions by comparing training-on-data rights, liability limits, and operational safeguards to determine acceptable terms versus escalated risks.

To check stacked-vendor accountability, the skill identifies flow-down failures between SaaS wrappers, orchestration layers, and model providers, ensuring obligations like incident notification and liability are contractually enforceable.

Yes, you can use this to perform a DPA versus AI addendum gap check, flagging where general privacy commitments exist but AI-specific governance terms are missing or inconsistent with your governance playbook.

To start reviewing an uploaded AI addendum, run the skill command with your vendor-terms-file; it confirms the document type, applies a standardized comparison framework, and outputs an escalation report.

A vendor AI contract review evaluates specific clauses including training on data, confidentiality of inputs, model change notice, output IP, liability, incident notification, human review rights, auditability, and data deletion obligations.