SEAT Assessment connector
The SEAT Security Awareness Maturity Framework, inside Claude. Answer 21 questions, get a maturity score for each of four pillars, map the result to one of 16 compliance and AI governance frameworks, and read your weakest outcomes against published research.
Connect
Add this as a custom connector in Claude (Settings, Connectors, Add custom connector). No account and no sign-in.
https://YOUR-DEPLOYMENT/mcp
Try it
- Run the SEAT assessment for my security awareness program.
- Map my SEAT results to NIS2 and list the missing evidence.
- What does the Verizon DBIR 2026 say about my weakest outcomes?
Tools
All eight tools are read-only. None writes, changes or deletes anything.
| Tool | What it does |
|---|---|
seat_get_overview | Pillars, maturity levels, and the published scoring rules. |
seat_get_questions | The 21 core questions with five answer options each, verbatim. Optional pillar filter. |
seat_get_facilitation_notes | Why a question exists, what strong looks like, and the trap that inflates scores. |
seat_score_assessment | Pillar and overall scores, levels, the five lowest questions with the next-level option. |
seat_list_frameworks | The 16 framework adapters and their ids. |
seat_map_to_framework | Met or gap per requirement, not-assessed rows, missing evidence, and a 0 to 5 alignment score. |
seat_list_research_sources | Eight published research reports mapped to SEAT outcomes, and the rules for using them. |
seat_get_research_source | One research source with the figures that may be cited, verbatim. |
What happens to your answers
Scoring and framework mapping receive the 21 numeric answers and nothing else: no names, no company details. They are processed in memory, not stored, and the server code does not log answer contents. The hosting platform keeps standard request metadata. Details are in the privacy policy.
If your organization cannot send answers to a third party, run the same assessment entirely inside your own environment from the public repository. This connector serves that repository's files and applies its published arithmetic.
Content and license
Questions, scoring rules, framework adapters and research summaries come unmodified from jason-hoenich/seat-assessment, licensed CC BY-ND 4.0. Research files summarize third-party reports with attribution and links to the originals. They do not replace the reports.
Hosted version
For saved results, recommendations and program plans, the hosted assessment is free at app.humanrisk.com.