kanonik — Claude Code MCP server: compliance with human sign-off
Wires your agent to the Kanonik service for policy and control work across ISO 27001, SOC 2, GDPR and HIPAA, with a verifier and human approval on every write.
claude mcp add --transport http kanonik https://app.kanonik.ai/mcpWhat kanonik does
Kanonik hands the agent a typed compliance model plus a record of its actions that cannot be quietly edited. A server-side Verifier checks the output, and each write waits behind a human approval gate. Nothing lands in your record without a single-use signed token and a person clicking approve. The agent cannot sign off on its own work.
The plugin is one file naming where the MCP server lives. No bundled code, no install script, no hook, no local process — everything runs at Kanonik. Your client connects to https://app.kanonik.ai/mcp and a browser window opens for sign-in on first use. A Kanonik account is required.
Sign-in uses OAuth 2.1 with Dynamic Client Registration and PKCE against https://auth.kanonik.ai; tokens stay in your client. Scopes are kanonik:read for reading the record, kanonik:propose for drafts, kanonik:commit for submitting them. Kanonik holds no credentials to your source systems, so if the agent pulls from GitHub, Jira or your identity provider it uses its own connections. Network access goes to one endpoint only.
What it helps with
- produce a Statement of Applicability against a loaded framework
- spot the gaps between framework requirements and what you already run
- attach evidence to every judgement about an internal control
- build an audit trail an assessor can verify on their own
- revise policies and procedures for ISO 27001:2022 or SOC 2
Example prompts for the agent
- Draft a Statement of Applicability for ISO 27001:2022 from my workspace
- Assess our internal controls and record the evidence behind each call
- Find where our GDPR coverage falls short of the framework
How to install kanonik
Platform CLI: remote server
claude mcp add --transport http kanonik https://app.kanonik.ai/mcpPlatform CLI: remote server
codex mcp add kanonik --url https://app.kanonik.ai/mcpConfig file
{
"mcpServers": {
"kanonik": {
"url": "https://app.kanonik.ai/mcp"
}
}
}Config file
{
"servers": {
"kanonik": {
"type": "http",
"url": "https://app.kanonik.ai/mcp"
}
}
}Platform CLI: remote server
gemini mcp add -t http kanonik https://app.kanonik.ai/mcpStep by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about kanonik
How do I install kanonik in Claude Code?
Run claude mcp add --transport http kanonik https://app.kanonik.ai/mcp. The first request opens a browser window so you can sign in to your Kanonik account.
What does kanonik do?
It gives the agent a compliance model and an action record, verifies the work server-side, and requires a human to approve each write before it is committed.
Which frameworks does kanonik cover?
ISO 27001:2022, SOC 2, GDPR, NIST CSF 2.0 and HIPAA. The framework in scope comes from your workspace, and nothing in the plugin is tied to one of them.
