Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add ski043/kinde-agent-skills --skill kinde-mcpgit clone --depth 1 https://github.com/ski043/kinde-agent-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ski043/kinde-agent-skills/kinde-mcp)<a href="https://agentmods.dev/skills/ski043/kinde-agent-skills/kinde-mcp"><img src="https://agentmods.dev/badge/skills/ski043/kinde-agent-skills/kinde-mcp/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ski043/kinde-agent-skills/kinde-mcp"><img src="https://agentmods.dev/badge/skills/ski043/kinde-agent-skills/kinde-mcp.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00074 | $0.02536 |
| Opus 5 | $0.00037 | $0.01268 |
| Sonnet 5 | $0.00015 | $0.00507 |
| Haiku 4.5 | $0.00007 | $0.00254 |
Grade A, and why
kinde-mcp scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 155 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Kinde MCP
Select exactly one product before any URL, key, or tool call. Never mix them.
Dated plan, client, and naming facts live in REFERENCES.md. Read it before relying on a maturity, entitlement, OAuth, or interpolation claim, and whenever official pages disagree.
This skill operates MCP; it does not redesign tenancy. For tenant operations exposed by an MCP connection, actively invoke kinde-b2b and consume its tenant-resolution/authorization rule. If that owner is unavailable, or the rule is absent, exposure is blocked.
Products
| Product | Endpoint | Auth | Owns |
|---|---|---|---|
| Management MCP | https://<business>.kinde.com/mcp |
MCP-enabled Environment API key only | A hosted subset of Kinde account data. Users and organizations are get-only. Creates exist only for the catalog/config tools in OPERATIONS.md. |
| MCP connections | Connection-specific URL from that connection's Quick start (path is not bare /mcp; host may be a custom domain) |
The live Quick start: documented user API key or an offered OAuth browser flow. Organization API keys are not documented client credentials. | The caller's registered OpenAPI operations, proxied to their API. Not Kinde account administration. |
Both servers are hosted. There is no official Kinde npm package or GitHub MCP server. Recheck current client support in REFERENCES.md and the live Quick start before connection setup. Prefer native remote HTTP when the selected client supports it.
Guardrails
- Discover, then call. The live tool list is evidence. Plan only tools that appear at runtime and exist in OPERATIONS.md (Management) or on the connection's Tools page (connections).
- Create is a mutation. Without authorization record
planned; with authorization call it and capture observed evidence. - Catalog create ≠ assignment.
CreateRole/CreatePermissionadd definitions. They do not assign a role to a user or attach a permission to a role. - Subscribers are leads.
GetSubscribers/CreateSubscriberare email/lead subscribers, not billing customers. - Expand is not billing admin. A docs sample that expands related billing data is a get-adjacent field. No billing tool is on the Management matrix.
- Credential class. An Environment API key is a global admin credential. A user API key binds to
user_id(org_codeis null). Consumekinde-b2bmachine-access rules; do not treat either key as the other product's credential. - Tenant connection. A user API key cannot establish a tenant. A tenant operation needs the
kinde-b2bcontract's named server-side tenant/membership resolution and authorization rule. OAuth follows tenant context established by the product API; do not invent claims.
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 155 lines · 74 tokens per session scan A 822190d3a36f
kinde-mcp is a skill published in the GitHub repository ski043/kinde-agent-skills (2 stars, last pushed 19d ago), licensed MIT. It adds 74 tokens to every session and 2,536 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
implement-saaskit-nextjs
Implements SaaSKit login in the Next.js App Router so an agent can add login, callback, session cookies, and logout. Use when the user wants SaaSKit Next.js or App Router auth. It does not write generic Node login (that's implement-saaskit) or write Django, FastAPI, or Flask auth (that's implement-saaskit-python).
implement-sso
Implements Modular SSO so an agent can add SAML/OIDC, IdP-initiated login, and the admin portal. Use when the user wants SSO, SAML, OIDC, Modular SSO, or IdP. It does not write SaaSKit login (that's implement-saaskit) or provision users from a directory (that's implement-scim).
expose-agentkit-mcp
Exposes AgentKit tools over MCP so a client can call them on a per-user instance URL. Use when the user wants to expose AgentKit over MCP, generate a per-user MCP URL, or connect LangChain via MCP. It does not list connectors (that's discover-connectors) or create a connected account and token in app code (that's…
implement-saaskit-python
Implements SaaSKit login in Django, FastAPI, or Flask so an agent can add login, callback, session cookies, and logout. Use when the user wants SaaSKit Django, FastAPI, or Flask auth. It does not write Node or Next.js login (that's implement-saaskit or implement-saaskit-nextjs).
implement-saaskit
Implements SaaSKit login so an agent can add a login route, a callback, session cookies, and logout. Use when the user wants to add SaaSKit login, a callback, or logout. It does not store, validate, refresh, or revoke a session after login (that's manage-saaskit-sessions) or write Next.js App Router auth (that's…
implement-scim
Implements SCIM so an agent can provision users and groups from a directory. Use when the user wants SCIM or to provision users. It does not write SaaSKit login (that's implement-saaskit) or add SSO (that's implement-sso).