sumsub-create-questionnaire — Claude Code skill for Sumsub forms
Turns a short spec into a Sumsub KYC questionnaire definition, pushes it to the API and hands back the id for level wiring.
npx skills add sumsub/agent-skills --skill sumsub-create-questionnaire -a claude-codeWhat sumsub-create-questionnaire does
The skill takes a compact list of sections and questions, expands it into the full JSON payload and sends it to Sumsub. New questionnaires go through POST /resources/api/agent/questionnaires, updates through PATCH on the same path, and GET /resources/api/agent/questionnaires/{id} reads one back. The work sits in build_questionnaire.py, post_questionnaire.sh, patch_questionnaire.sh and get_questionnaire.sh; the spec arrives on stdin, so you never hand-write the localized payload.
Before anything else it checks tenant entitlements with sumsub-check-permissions: QUESTIONNAIRE has to appear in allowedChecks. Miss it and the skill stops, pointing you at your CSM or Sumsub support. Auth uses an App Token plus secret, and only sandbox pairs are accepted — the helper script rejects tokens that do not start with sbx:. A production token gets refused, so this is not the route for a live tenant.
POST and PATCH are strict now. POST returns 409 CONFLICT when the id is taken, PATCH returns 404 NOT_FOUND for an unknown one. Unsure whether the questionnaire exists? Call get_questionnaire.sh first. The response carries id and clientId, which the skill turns into a sandbox dashboard link.
What it helps with
- build a KYC questionnaire definition from a list of questions
- update an existing Sumsub questionnaire by its id
- read a questionnaire back and check what landed
- wire conditional questions to real section and item ids
- get a sandbox link to the questionnaire for review
Example prompts for the agent
- create a Sumsub questionnaire with a source-of-funds section and a salary question
- update the source-of-funds questionnaire and add an other-source question
- read the source-of-funds questionnaire and show me what is in it now
How to install sumsub-create-questionnaire
1Skills CLI (Vercel)
npx skills add sumsub/agent-skills --skill sumsub-create-questionnaire -a claude-code2Manually: put the files in place
~/.claude/skills/sumsub-create-questionnaire/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/sumsub-create-questionnaire/
1Skills CLI (Vercel)
npx skills add sumsub/agent-skills --skill sumsub-create-questionnaire -a codex2Manually: put the files in place
.agents/skills/sumsub-create-questionnaire/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add sumsub/agent-skills --skill sumsub-create-questionnaire -a cursor2Manually: put the files in place
.cursor/skills/sumsub-create-questionnaire/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install sumsub/agent-skills sumsub-create-questionnaire2Manually: put the files in place
.github/skills/sumsub-create-questionnaire/SKILL.md1Skills CLI (Vercel)
npx skills add sumsub/agent-skills --skill sumsub-create-questionnaire -a gemini-cli2Platform CLI
gemini skills install https://github.com/sumsub/agent-skills --path skills/sumsub-create-questionnaireManually: put the files in place
.devin/skills/sumsub-create-questionnaire/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add sumsub/agent-skills --skill sumsub-create-questionnaire -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about sumsub-create-questionnaire
How do I install sumsub-create-questionnaire in Claude?
Run npx skills add sumsub/agent-skills --skill sumsub-create-questionnaire -a claude-code. The skill lands in ~/.claude/skills/sumsub-create-questionnaire/SKILL.md.
What does sumsub-create-questionnaire do?
It builds a Sumsub questionnaire definition from a compact spec and sends it to the API — POST for a new one, PATCH for an existing one. The response gives you the id and a sandbox link.
What permissions and tokens does the skill need?
The QUESTIONNAIRE key must be present in allowedChecks; the skill verifies this through sumsub-check-permissions and stops if it is missing. Only sandbox tokens work — the script rejects anything that does not start with sbx:.