Is xmcp safe?
- Node.js shell/command execution
- Node.js network egress
- Delegated authentication (OAuth 2.0 / OIDC)
What to do: Nothing here argues against installing it. Grant the capabilities it lists only if you expect the tool to need them.
xmcp is an npm package analyzed by SkillTotal's deterministic static scanner. The scan found no malicious indicators. It can: delegated authentication, mcp tools detected, network egress and shell execution — capabilities are what the code can do, not a verdict on intent. Risk score 0/100 (low).
xmcp 1.9.0
Get notified when this component changes
Change alerts aren't live yet. Leave your email and we'll let you know when they launch. We use it only for this and delete it on request.
Automated static-analysis result. It can contain false positives and false negatives, and is not a claim about the intent of xmcp's authors. Report a false positive.
Behavioral traits
How this component maps to the CSA agentic threat model. Descriptive — it never affects the risk score.
Findings (4)
The component can run operating-system commands or spawn processes.
(()=>{var t,r,o={6507(t,r,o){"use strict";let a=o(5317),i=o(8388),n=o(442);function s(t,r,o){let s=i(t,r,o),l=a.spawn(s.command,s.args,s.options);return n.hookChildProcess(l,s),l}t.exports=s,t.exports.spawn=s,t.exports.sync=function(t,r,o){ …!function(t,r){if("object"==typeof exports&&"object"==typeof module)module.exports=r(require("zod/v3"));else if("function"==typeof define&&define.amd)define(["zod/v3"],r);else{var a="object"==typeof exports?r(require("zod/v3")):r(t["zod/v3" …Why it matters: Powerful and often legitimate — confirm the commands aren't built from untrusted input.
Fix: Confirm the command and its arguments are fully controlled and not derived from untrusted input; prefer execFile with an argument array.
The component makes outbound network requests.
`),k=!0;try{h?.()}catch(t){i?.(t instanceof Error?t:Error(String(t)))}void 0!==f&&clearInterval(f),g?.(),a.delete($);try{m.close()}catch{}}},T=new ReadableStream({start(t){m=t;let i=dc({method:"notifications/subscriptions/acknowledged",para …Why it matters: Usually legitimate, but confirm the destinations are expected and no sensitive data leaves.
Fix: Confirm the destination hosts are expected and that no sensitive data is sent off-host.
An OAuth 2.0 / OpenID Connect delegated-authentication flow was detected (authorization-code / refresh-token / token-exchange grant, an OIDC authorize/discovery endpoint or id_token, or a delegation library). Tools authenticate with the end user's delegated, scoped credentials rather than a long-lived embedded service credential. (5 occurrence(s) shown as evidence).
Set the \`cycles\` parameter to \`"ref"\` to resolve cyclical schemas with defs.`)}for(let o of t.seen.entries()){let a=o[1];if(r===o[0]){i(o);continue}if(t.external){let a=t.external.registry.get(o[0])?.id;if(r!==o[0]&&a){i(o);continue}}if …`)&&-1===t.indexOf("\r")){s.push(t),l+=t.length,m();return}s.push(t);let r=s.join("");s.length=0,l=0;let o=y(r);""!==o&&(s.push(o),l=o.length),m()},reset:function(t={}){if(t.consume&&s.length>0){let t=s.join("");g(t,0,t.length)}c=!0,u=void …Set the \`cycles\` parameter to \`"ref"\` to resolve cyclical schemas with defs.`)}for(let a of t.seen.entries()){let o=a[1];if(r===a[0]){i(a);continue}if(t.external){let o=t.external.registry.get(a[0])?.id;if(r!==a[0]&&o){i(a);continue}}if …`)&&-1===t.indexOf("\r")){s.push(t),l+=t.length,m();return}s.push(t);let r=s.join("");s.length=0,l=0;let a=y(r);""!==a&&(s.push(a),l=a.length),m()},reset:function(t={}){if(t.consume&&s.length>0){let t=s.join("");g(t,0,t.length)}c=!0,u=void …Set the \`cycles\` parameter to \`"ref"\` to resolve cyclical schemas with defs.`)}for(let i of t.seen.entries()){let n=i[1];if(r===i[0]){o(i);continue}if(t.external){let n=t.external.registry.get(i[0])?.id;if(r!==i[0]&&n){o(i);continue}}if …Fix: Delegated auth is a lower-blast-radius execution context than an embedded static credential. Confirm the requested scopes are minimal and that tokens are never logged or forwarded off-host.
An MCP tool surface (manifest or tool definitions) was found.
}`)}}return c}}(p,S,s,d.name);let w={title:d.annotations?.title,description:d.description,inputSchema:m6(g),outputSchema:s?m6(s,{strict:!0}):void 0,annotations:d.annotations,icons:d.icons,_meta:b};t.registerTool(d.name,w,(a={name:d.name,des …Why it matters: Just context — review which tools it offers and their permissions.
Fix: Review the declared MCP tools and their permissions.
How attackers abuse these capabilities
Interactive labs on the attack class behind the rules above. They show the technique, not anything found in xmcp.
The free Claude Code plugin scans each npm, PyPI or MCP package with this report's open-source engine before the install command runs. It runs on your machine and needs no account.
Check your own component
Run the same evidence-backed scan on any MCP server, agent skill, or package.
Scan your own componentHow we determine this: deterministic static analysis (regex + AST), evidence-anchored, no code execution. Methodology →