Atlas / MCP servers / 1clawai / @1claw/sdk (v0.61.0)

@1claw/sdk (v0.61.0)CAUTION

mcp/1clawai/1claw-sdk

TypeScript SDK for 1Claw Vault — secret management for AI agents

Verdict
CAUTION
Grade
B
Trust score
85 /100
Exposed tools
11 7r · 3w · 1d
Transport
—
License
MIT
Stars
2
01

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

⭐ Star [1clawAI/agent-templates](https://github.com/1clawAI/agent-templates) — ready-to-run agent templates wired to 1Claw. It is our single starred repo.

TypeScript/JavaScript client for the 1Claw Vault API.

This is the official SDK for Node.js, Next.js, and browser apps that call 1Claw over HTTP. It covers vaults, secrets, agents, policies, treasury, Intents API signing, execution bindings, platform apps, billing, and x402 payments. Types are generated from the OpenAPI spec, so request shapes stay in sync with the API.

Agent keys auto-exchange for JWTs and refresh before expiry. If you're wiring an AI agent in Cursor or Claude Desktop, you probably want @1claw/mcp instead. If you're building a web app or backend service in TypeScript, start here.

Install

npm install @1claw/sdk

Note: This package is ESM-only. Use it in an ESM context (e.g. "type": "module" in package.json, or .mjs files). Next.js and other bundlers handle ESM natively. Running scripts with plain node or tsx may require an ESM setup to avoid ERR_PACKAGE_PATH_NOT_EXPORTED.

Quick Start

import { createClient } from "@1claw/sdk";

const client = createClient({
baseUrl: "https://api.1claw.co",
apiKey: "ocv_...", // auto-exchanges for a JWT
});

// List vaults
const { data } = await client.vault.list();
console.log(data?.vaults);

// Store a secret
await client.secrets.set("vault-id", "OPENAI_KEY", "sk-...", {
type: "api_key",
});

// Retrieve a secret.
// Every call returns { data, error }. `data` is null whenever `error` is set,
// so `data?.value` alone turns a 403 into a silent undefined — check `error`.
const { data, error, meta } = await client.secrets.get("vault-id", "OPENAI_KEY");
if (error) {
console.error(meta?.status, error.type, error.message, meta?.requestId);
} else {
console.log(data?.value);
}

API contract: This SDK is built from

Read from source at commit 798d243a6ff9OBSERVED · 2026-10-12
02

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.

claude-code
claude mcp add sdk -- npx -y @1claw/[email protected]
claude-desktop
{
  "mcpServers": {
    "sdk": {
      "command": "npx",
      "args": [
        "-y",
        "@1claw/[email protected]"
      ]
    }
  }
}
03

Exposed tools (11)

7 read · 3 write · 1 destructive. Blast radius: 1 tool can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.

ToolRiskDescription
1claw_check_approval_statusreadCheck the current status of a human approval request.
1claw_create_vaultwriteCreate a new encrypted vault in 1Claw.
1claw_delete_secretdestructivePermanently delete a secret from a 1Claw vault.
1claw_get_secretreadFetch a decrypted secret from a 1Claw vault.
1claw_list_secret_keysreadList all secret keys in a vault WITHOUT revealing their values.
1claw_list_vaultsreadList all vaults accessible to the current identity.
1claw_pay_and_fetchwritePay the x402 micropayment fee and retrieve a secret.
1claw_request_approvalreadFormally request human approval to access a gated secret.
1claw_set_secretwriteStore or update a secret in a 1Claw vault.
1claw_share_secretreadShare a secret with another person by email.
prodreadtest
04

Trust audit

CAUTIONgrade B · trust 85/100 Install with care. The audit found things worth knowing before you trust its output.

LayerWhat it checksResult
L0Provenance & inventoryPASS
L1Static analysis of the codeWARN
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfaceWARN
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
declared (1 observation(s))
Network
none-observed
Shell
none-observed
Dependencies
not all pinned
Secrets in source
found

Findings (12)

MEDIUMHard-coded secrets · secret.generic · CWE-798, CWE-321
src/__tests__/errors.test.ts:206
apiKey: "ocv_notarealkey0000000000",
MEDIUMHard-coded secrets · secret.generic · CWE-798, CWE-321
src/__tests__/errors.test.ts:224
apiKey: "ocv_notarealkey0000000000",
MEDIUMFilesystem / path · mcp.destructive_tools · CWE-22, CWE-59
1claw_delete_secret
Why it matters. 1 tool(s) can delete or overwrite
Fix. prefer a read-only mode or scoped tokens; the page states the blast radius
LOWInventory / provenance · inv.hidden_file · CWE-1104
.claudeignore
.claudeignore
Why it matters. hidden member outside the usual dotfiles
Fix. review its purpose
LOWInventory / provenance · inv.hidden_file · CWE-1104
.cursorignore
.cursorignore
Why it matters. hidden member outside the usual dotfiles
Fix. review its purpose
LOWObfuscation / stealth · obf.decode_call · CWE-506, CWE-94
src/cmek.ts:138
const binary = atob(b64);
LOWObfuscation / stealth · obf.decode_call · CWE-506, CWE-94
src/core/http.ts:113
atob(parts[1].replace(/-/g, "+").replace(/_/g, "/")),
LOWSupply chain · supply.unpinned · CWE-829, CWE-1357
package.json
@types/node, openapi-typescript, typescript, vitest
Why it matters. 4 dependency range(s) float
Fix. pin exact versions or ship a lockfile
LOWPrompt injection · prompt.transfer_instruction · CWE-94, CWE-1427
README.md:632
Vaults can optionally use multi-party computation (MPC) for secret splitting. When MPC is enabled, the server holds one share and the client holds another — neither party can reconstruct the secret al
Why it matters. an instruction to move sensitive data to an outside destination
Fix. remove; a skill never needs the user's secrets off the machine
LOWPrompt injection · prompt.transfer_instruction · CWE-94, CWE-1427
README.md:652
When MPC is enabled, `set` responses include a `client_share` field that must be stored client-side. The `get` method accepts an optional `client_share` parameter to reconstruct the full secret. Witho
Why it matters. an instruction to move sensitive data to an outside destination
Fix. remove; a skill never needs the user's secrets off the machine
LOWPrompt injection · prompt.transfer_instruction · CWE-94, CWE-1427
README.md:699
const proof = await dpop.generateProof("POST", "https://api.1claw.co/v1/auth/agent-token");
Why it matters. an instruction to move sensitive data to an outside destination
Fix. remove; a skill never needs the user's secrets off the machine
INFOInventory / provenance · inv.oversize · CWE-1104
src/generated/api-types.ts
src/generated/api-types.ts
Why it matters. 1220170 bytes not read

Gates applied: no_behavioural_pass.

Audited 2026-10-12 · audit v0.4.1 · source sha 798d243a6ff9full audit observations/trust-audit/mcp-server/1clawai__1claw-sdk.json · Report an issue / request a re-scan
05

Audit history

Every audit this server has had. A grade with a past is a grade somebody is still checking.

DateSourceVerdictGradeScoreChange
2026-10-12798d243a6ff9CAUTIONB85first audit
06

Questions

What is the @1claw/sdk MCP server?

TypeScript SDK for 1Claw Vault — secret management for AI agents

What tools does @1claw/sdk expose?

11 in total: 7 read-only, 3 that write, and 1 that can delete or overwrite (1claw_delete_secret). Every one is listed on this page with its risk.

Is @1claw/sdk safe to connect to an agent?

With care. The audit graded it B (85/100) and found 12 things worth knowing before you trust this server, listed below with the exact line each was found on. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.

What credentials does @1claw/sdk need?

No credential environment variables were found in its source, so it appears to need none.

How current is this page?

The grade is for one exact copy of the source (798d243a6ff9), read on 2026-10-12. The repository is watched and re-audited when it changes.

Advertisement