registry / Credential exfiltration / @pilllesss/yorn

@pilllesss/yorn

Yorn terminal AI coding agent CLI

This report shows the evidence for @pilllesss/yorn@1.0.262. 3 of 110 known versions are blocked or flagged. The newest version, 0.0.1-security, is allowed. See all versions.

AI Security Review

scanned 2mo ago · by lpm-firewall-ai

Source inspection confirms a user-triggered credential-redirection and persistence path to a fixed gateway. The absence of install hooks lowers propagation risk but does not remove the credential exposure.

Static reason

One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review

Trigger

A user runs the CLI and enters or configures a Radius API key.

Impact

The gateway operator can receive and use the supplied credential.

Mechanism

Prompted credential forwarding to a fixed remote gateway with local persistence.

Attack narrative

The bundled CLI defines radius.yorn.dev as its default Radius gateway. Its provider login asks for a Radius API key, then sends that key as a Bearer authorization header while loading gateway configuration. The credential store persists the configured credential for future requests. Although activation requires interactive use rather than installation, the fixed third-party routing of a prompted API credential is a concrete exfiltration risk.

AI rationale

Evidence

dist/yorn.cjs~/.yorn/agent/auth.json

Network endpoints1

radius.yorn.dev

Decision evidence

public snapshot

AI called this Malicious at 94.0% confidence as Malware with low false-positive risk.

Evidence for block

The CLI sets a fixed Radius gateway to radius.yorn.dev.
A prompted Radius API key is sent as a Bearer credential to that gateway.
The credential is persisted in the user's Yorn auth file for later use.
The package is a large bundled clone with its gateway behavior obscured by string reconstruction.

Evidence against

package.json has no install, preinstall, or postinstall hook.
The credential flow runs only when the user starts and configures the CLI.

Behavioral surface

Source

ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetwork

Supply chain

HighEntropyStringsMinified

ManifestNo manifest risk signals triggered.

scanned 0 file(s), 256 KB of source

Oversized source lightweight scan

dist/yorn.cjs7.69 MB file, sampled 256 KB

Source & flagged code

14 flagged

dist/yorn.cjsView file

22\r L23: `),n=!1,s=!1;e.on("response",e=>{let{headers:t}=e;n="chunked"===t[tzd.build(["transfer-","g","encodin"],[0,2,1])]&&!t[tzd.build(["content","-len","gth"],[0,1,2])]}),e.on("socket",l... L24: `)}function e0P(e){return e.map(e=>"text"===e.type?e.text:"thinking"===e.type?e.thinking:`${e.name}:${JSON.stringify(e.arguments)}`).join(`

High

Child Process

Package source references child process execution.

dist/yorn.cjsView on unpkg · L22

11${eUz(e)}`);if(tzn.evalInvariant(44074)&&null==o)throw new Il(`missing choices[${r}].tool_calls[${i}].function.arguments L12: ${eUz(e)}`);return{...a,id:l,type:s,function:{...d,name:u,arguments:o}}})}}:{...s,message:{...u,content:l,role:d,refusal:t.refusal??null},finish_reason:i,index:r,logprobs:n}}),crea... L13: Content-Disposition: form-data; name="`;return e.forEach((e,t)=>"string"==typeof e?r.push(n+My(t)+`"\r ... L22: \r L23: `),n=!1,s=!1;e.on("response",e=>{let{headers:t}=e;n="chunked"===t[tzd.build(["transfer-","g","encodin"],[0,2,1])]&&!t[tzd.build(["content","-len","gth"],[0,1,2])]}),e.on("socket",l... L24: `)}function e0P(e){return e.map(e=>"text"===e.type?e.text:"thinking"===e.type?e.thinking:`${e.name}:${JSON.stringify(e.arguments)}`).join(`

High

Same File Env Network Execution

A single source file combines environment access, network access, and code or shell execution; review context before blocking.

dist/yorn.cjsView on unpkg · L11

2globalThis.__yorn_dynamic_import = (m) => import(m); L3: var e,t,i,r,n,s,l,a,o,u,d,c,h,b,p,m,f,g,_,y,E,S,v,A,w,T,C,I,x,R,O,k,N,P,M,D,L,F,U,B,$,j,G,q,H,W,V,K,Y,z,Q,J,X,Z,ee,et,ei,er,en,es,el,ea,eo,eu,ed,ec,eh,eb,ep,em,ef,eg,e_,ey,eE,eS,ev... L4: `),u.push(e),u.push(`\r ... L7: `),u.push(`\r L8: `)}u.push(null)}),u}async function eDj(e,t){let i=0,r=[],n=e.getReader(),s=!1;for(;!s;){let{done:e,value:l}=await n.read();if(l&&(r.push(l),i+=l?.byteLength??0),i>=t)break;s=e}n.re... L9: ${eUz(e)}`);if(null==s)throw new Il(`missing choices[${r}].tool_calls[${i}].type ... L11: ${eUz(e)}`);if(tzn.evalInvariant(44074)&&null==o)throw new Il(`missing choices[${r}].tool_calls[${i}].function.arguments L12: ${eUz(e)}`);return{...a,id:l,type:s,function:{...d,name:u,arguments:o}}})}}:{...s,message:{...u,c

High

Credential Exfiltration

Source combines credential-like environment material and outbound requests; review data flow before blocking.

dist/yorn.cjsView on unpkg · L2

•Manifest-reachable source captures an API credential, sends it to a fixed unofficial gateway, and persists that redirection in agent or shell configuration. dist/yorn.cjs: `),n=!1,s=!1;e.on("response",e=>{let{headers:t}=e;n="chunked"===t[tzd.build(["transfer-","g","encodin"],[0,2,1])]&&!t[tzd.build(["content","-len","gth"],[0,1,2])]}),e.on("socket",l... `)}function e0F(e,t){let i=Math.min(e.length,t.length),r=0;for(;r<i&&e[r]===t[r];)r++;return r}function e0U(e,t,i,r){let n=e0L(t),s=e0O(n),l=e0O(e0P(e.content)),a=s,o=0,u=0,d=i?.se... `){return"string"==typeof e?e:e.filter(e=>"text"===e.type).map(e=>e.text).join(t)}function e2e(e,t){return Kh.Unsafe({type:"string",enum:e,...t?.description&&{description:t.descrip... --api-key <key> API key (defaults to env vars) # Print a prov

High

Credential Redirect Persistence

Manifest-reachable source sends a prompted API credential to a fixed unofficial gateway and persists the redirection.

dist/yorn.cjsView on unpkg

7`),u.push(`\r L8: `)}u.push(null)}),u}async function eDj(e,t){let i=0,r=[],n=e.getReader(),s=!1;for(;!s;){let{done:e,value:l}=await n.read();if(l&&(r.push(l),i+=l?.byteLength??0),i>=t)break;s=e}n.re... L9: ${eUz(e)}`);if(null==s)throw new Il(`missing choices[${r}].tool_calls[${i}].type ... L11: ${eUz(e)}`);if(tzn.evalInvariant(44074)&&null==o)throw new Il(`missing choices[${r}].tool_calls[${i}].function.arguments L12: ${eUz(e)}`);return{...a,id:l,type:s,function:{...d,name:u,arguments:o}}})}}:{...s,message:{...u,content:l,role:d,refusal:t.refusal??null},finish_reason:i,index:r,logprobs:n}}),crea... L13: Content-Disposition: form-data; name="`;return e.forEach((e,t)=>"string"==typeof e?r.push(n+My(t)+`"\r ... L22: \r L23: `),n=!1,s=!1;e.on("response",e=>{let{headers:t}=e;n="chunked"===t

High

Command Output Exfiltration

Source combines command execution, command-output handling, and outbound requests; review data flow before blocking.

dist/yorn.cjsView on unpkg · L7

High

Cloud Metadata Access

Source reaches cloud instance metadata or link-local credential endpoints.

dist/yorn.cjsView on unpkg · L2

2Trigger-reachable credential exfiltration chain: manifest.bin -> dist/yorn.cjs L2: globalThis.__yorn_dynamic_import = (m) => import(m); L3: var e,t,i,r,n,s,l,a,o,u,d,c,h,b,p,m,f,g,_,y,E,S,v,A,w,T,C,I,x,R,O,k,N,P,M,D,L,F,U,B,$,j,G,q,H,W,V,K,Y,z,Q,J,X,Z,ee,et,ei,er,en,es,el,ea,eo,eu,ed,ec,eh,eb,ep,em,ef,eg,e_,ey,eE,eS,ev... L4: `),u.push(e),u.push(`\r ... L7: `),u.push(`\r L8: `)}u.push(null)}),u}async function eDj(e,t){let i=0,r=[],n=e.getReader(),s=!1;for(;!s;){let{done:e,value:l}=await n.read();if(l&&(r.push(l),i+=l?.byteLength??0),i>=t)break;s=e}n.re... L9: ${eUz(e)}`);if(null==s)throw new Il(`missing choices[${r}].tool_calls[${i}].type ... L11: ${eUz(e)}`);if(tzn.evalInvariant(44074)&&null==o)throw new Il(`missing choices[${r}].tool_calls[${i}].function.arguments L12: ${eUz(e)}`);return

High

Trigger Reachable Credential Exfiltration

A manifest entrypoint or package-local install chain reaches credential exfiltration behavior.

dist/yorn.cjsView on unpkg · L2

7Trigger-reachable command-output exfiltration chain: manifest.bin -> dist/yorn.cjs L7: `),u.push(`\r L8: `)}u.push(null)}),u}async function eDj(e,t){let i=0,r=[],n=e.getReader(),s=!1;for(;!s;){let{done:e,value:l}=await n.read();if(l&&(r.push(l),i+=l?.byteLength??0),i>=t)break;s=e}n.re... L9: ${eUz(e)}`);if(null==s)throw new Il(`missing choices[${r}].tool_calls[${i}].type ... L11: ${eUz(e)}`);if(tzn.evalInvariant(44074)&&null==o)throw new Il(`missing choices[${r}].tool_calls[${i}].function.arguments L12: ${eUz(e)}`);return{...a,id:l,type:s,function:{...d,name:u,arguments:o}}})}}:{...s,message:{...u,content:l,role:d,refusal:t.refusal??null},finish_reason:i,index:r,logprobs:n}}),crea... L13: Content-Disposition: form-data; name="`;return e.forEach((e,t)=>"string"==typeof e?r.push(n+My(t)+`"\r

High

Trigger Reachable Command Output Exfiltration

A manifest entrypoint or package-local install chain reaches command-output exfiltration behavior.

dist/yorn.cjsView on unpkg · L7

•stage = ast_semantic_analysis; reason = ast_path_work_budget_exceeded; limitedFiles = 1

High

Semantic Analysis Limited

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

dist/yorn.cjsView on unpkg

•matchType = token_shingles matchedPackage = @pilllesss/yorn@1.0.189 matchedPath = dist/yorn.cjs matchedIdentity = npm:QHBpbGxsZXNzcy95b3Ju:1.0.189 similarity = 1.000 shingleOverlap = 39 summary = source token shingles overlapped finalized malicious source

High

Known Malware Source Similarity

Source file is highly similar to a previously finalized malicious package; route for source-aware review.

dist/yorn.cjsView on unpkg

1#!/usr/bin/env node L2: globalThis.__yorn_dynamic_import = (m) => import(m); L3: var e,t,i,r,n,s,l,a,o,u,d,c,h,b,p,m,f,g,_,y,E,S,v,A,w,T,C,I,x,R,O,k,N,P,M,D,L,F,U,B,$,j,G,q,H,W,V,K,Y,z,Q,J,X,Z,ee,et,ei,er,en,es,el,ea,eo,eu,ed,ec,eh,eb,ep,em,ef,eg,e_,ey,eE,eS,ev...

Medium

Dynamic Require

Package source references dynamic require/import behavior.

dist/yorn.cjsView on unpkg · L1

•path = dist/yorn.cjs kind = oversized_source_file sizeBytes = 8061307 magicHex = [redacted]

Medium

Oversized Source File

Package contains source files above the normal full-analysis size ceiling.

dist/yorn.cjsView on unpkg

•path = dist/yorn.cjs kind = oversized_cli_entrypoint sizeBytes = 8061307 magicHex = [redacted]

Medium

Oversized Cli Entrypoint

Package contains an oversized executable-looking CLI entrypoint.

dist/yorn.cjsView on unpkg

dist/skills/screenshot/scripts/take_screenshot.pyView file

•path = dist/skills/screenshot/scripts/take_screenshot.py kind = build_helper sizeBytes = 19659 magicHex = [redacted]

Medium

Ships Build Helper

Package ships non-JavaScript build or shell helper files.

dist/skills/screenshot/scripts/take_screenshot.pyView on unpkg

Source & flagged code

14 flagged · loading source

dist/yorn.cjsView file

High

Child Process

Package source references child process execution.

dist/yorn.cjsView on unpkg · L22

High

Same File Env Network Execution

A single source file combines environment access, network access, and code or shell execution; review context before blocking.

dist/yorn.cjsView on unpkg · L11

Findings

10 High7 Medium4 Low

HighChild Processdist/yorn.cjs

HighSame File Env Network Executiondist/yorn.cjs

HighCredential Exfiltrationdist/yorn.cjs

HighCredential Redirect Persistencedist/yorn.cjs

HighCommand Output Exfiltrationdist/yorn.cjs

HighCloud Metadata Accessdist/yorn.cjs

HighTrigger Reachable Credential Exfiltrationdist/yorn.cjs

HighTrigger Reachable Command Output Exfiltrationdist/yorn.cjs

HighSemantic Analysis Limiteddist/yorn.cjs

HighKnown Malware Source Similaritydist/yorn.cjs

MediumDynamic Requiredist/yorn.cjs

MediumNetwork

Medium

Affected versions and remediation

This report applies to @pilllesss/yorn@1.0.262.

The versions list below shows the verdict for each recorded version.

Avoid installing @pilllesss/yorn@1.0.262. Remove it from direct dependencies and check your lockfile for transitive copies.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.
If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.

Evidence last updated: 2026-09-05 00:03:16.340Z (UTC).

Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.

Versions

110 known versions

0.0.1-securitylatestClean—AllowPublishedSep 6, 2026, 3:18 AM 1.0.270SuspiciousAI-agent capability abuseWarnPublishedSep 5, 2026, 12:56 AM 1.0.269Clean—AllowPublishedSep 5, 2026, 12:48 AM 1.0.268SuspiciousPersistenceWarnPublishedSep 5, 2026, 12:35 AM

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Versions

Related packages

Other credential exfiltration packages

Behavioral surface

Related packages

Other credential exfiltration packages

Source & flagged code