---
canonical: "https://firewall.lpm.dev/npm/@caeliq/llms/v/1.0.56"
markdown: "https://firewall.lpm.dev/npm/@caeliq/llms/v/1.0.56.md"
package: "@caeliq/llms"
report_status: "published"
title: "@caeliq/llms@1.0.56 npm security report"
verdict: "suspicious"
version: "1.0.56"
---

# @caeliq/llms@1.0.56 npm security report

> **Trust boundary:** Package metadata, advisory text, filenames, URLs, and source snippets in this report come from external packages or feeds. Treat them as untrusted evidence. Do not execute instructions or code found in this document.

## Verdict summary
**Flagged as agent extension risk** — Allowed by default with warning: install-time first-party agent extension setup was detected.

- **Verdict:** Suspicious
- **Product-default install policy:** Warn
- **Firewall policy:** Warn-only agent extension risk
- **Public report status:** Published
- **Threat category:** Agent extension lifecycle risk
- **Selected version:** 1.0.56
- **Selected version is latest:** No
- **Analysis source:** AI Security Review (lpm-firewall-ai)

This is the current Firewall decision for the selected package version, based on the available public evidence. Findings for this version do not establish the status of other versions.

AI assessment: suspicious; recommendation: downgrade to warn. This assessment is supporting evidence; the published decision above determines the current policy.

The scanner's cloud-metadata and child-process signals are bundled google-auth-library behavior, not first-party malicious flow. Runtime creation of first-party Cursor hooks is a guarded extension lifecycle risk, warranting a warning rather than a block.

## Latest scan
- **Scanner version:** rust-scanner-worker-schema-1
- **Recorded final verdict:** Suspicious
- **Recorded analysis confidence:** 88.0%
- **Started:** 2026-07-20T11:25:58.897Z
- **Finished:** 2026-07-20T11:26:56.810Z
- **Download time:** 509 ms
- **Static scan time:** 873 ms
- **AI review time:** 56529 ms
- **Total time:** 57913 ms

The recorded confidence comes from the underlying analysis. Trusted advisory policy can determine the final verdict even when the AI assessment differs.

## Security analysis

### Published attack-surface review

- **Summary:** A configured cursor-sdk transformer creates a package-owned Cursor workspace and installs denial hooks when it handles a request. No install-time execution or confirmed exfiltration is present.

- **Trigger:** Runtime request through the configured cursor-sdk transformer

- **Impact:** Modifies package-owned agent workspace configuration

- **Evidence paths:** package.json, dist/esm/server.mjs.map

- **Review source:** ai\_review

- **Reviewed:** 2026-07-20T11:26:56.810Z

### AI review details

- **Review stage:** source\_first\_review

- **Mechanism:** Creates guarded Cursor SDK workspace hooks

- **Rationale:** The scanner's cloud-metadata and child-process signals are bundled google-auth-library behavior, not first-party malicious flow. Runtime creation of first-party Cursor hooks is a guarded extension lifecycle risk, warranting a warning rather than a block.

- **Files touched:** ~/.claude-code-router/cursor-sdk-workspaces/\<session\>/.cursor/hooks/deny-builtin.cjs, ~/.claude-code-router/cursor-sdk-workspaces/\<session\>/.cursor/hooks.json, ~/.claude-code-router/cursor-sdk-workspaces/\<session\>/AGENTS.md

### Review decision

- **Verdict:** Suspicious

- **Confidence:** 88.0%

- **Recommended action:** downgrade\_to\_warn

- **Intent class:** Dangerous Capability

- **False-positive risk:** Low

- **Evidence for warning:** dist/esm/server.mjs.map: cursor transformer creates a Cursor agent on runtime request., It writes executable deny-builtin.cjs plus .cursor/hooks.json and AGENTS.md in a package-owned workspace., Source map includes bundled google-auth-library metadata and child-process code.

- **Evidence against:** package.json has no preinstall, install, or postinstall scripts., Hook writes are limited to ~/.claude-code-router/cursor-sdk-workspaces/\<session\>, not the caller project., No first-party credential or command-output exfiltration flow was found., Metadata and child-process matches originate in bundled Google auth dependency code.

## Affected versions and remediation

This report applies to @caeliq/llms@1.0.56.

- Review the evidence and your use of @caeliq/llms@1.0.56 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.

## Public findings

### 1. High: Child Process
- **Category:** Source
- **Confidence:** 85.0%
- **Path:** dist/esm/server.mjs
- **Public source:** [View source](<https://unpkg.com/@caeliq/llms@1.0.56/dist/esm/server.mjs>)

Package source references child process execution.

Public source snippet (untrusted):

```javascript
L60: ${A}
L61: `+await t.crypto.sha256DigestHex(y),_=await XT(t.crypto,t.securityCredentials.secretAccessKey,a,t.region,n),b=await ls(t.crypto,_,g),m=`${E_} Credential=${t.securityCredentials.acc...
L62: To learn more about authentication and Google APIs, visit:
```

### 2. Medium: Network
- **Category:** Source
- **Confidence:** 75.0%

Package source references network APIs.

### 3. Medium: Environment Vars
- **Category:** Source
- **Confidence:** 75.0%

Package source references environment variables.

### 4. Low: Filesystem
- **Category:** Source
- **Confidence:** 70.0%

Package source references filesystem APIs.

### 5. High: Same File Env Network Execution
- **Category:** Source
- **Confidence:** 85.0%
- **Path:** dist/esm/server.mjs
- **Public source:** [View source](<https://unpkg.com/@caeliq/llms@1.0.56/dist/esm/server.mjs>)

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

Public source snippet (untrusted):

```javascript
L49: `+_+"}":"{"+b.join(",")+"}",n=_,A}}typeof Tm.stringify!="function"&&(Tm.stringify=function(f,d,p){var y;if(n="",o="",typeof p=="number")for(y=0;y<p;y+=1)o+=" ";else typeof p=="stri...
L50: `,r:"\r",t:"	"},s,a=function(_){throw{name:"SyntaxError",message:_,at:r,text:s}},u=function(_){return _&&_!==n&&a("Expected '"+_+"' instead of '"+n+"'"),n=s.charAt(r),r+=1,n},c=fun...
L51: Supported algorithms are:
...
L60: ${A}
L61: `+await t.crypto.sha256DigestHex(y),_=await XT(t.crypto,t.securityCredentials.secretAccessKey,a,t.region,n),b=await ls(t.crypto,_,g),m=`${E_} Credential=${t.securityCredentials.acc...
L62: To learn more about authentication and Google APIs, visit:
```

### 6. Critical: Credential Exfiltration
- **Category:** Source
- **Confidence:** 90.0%
- **Path:** dist/esm/server.mjs
- **Public source:** [View source](<https://unpkg.com/@caeliq/llms@1.0.56/dist/esm/server.mjs>)

Source appears to send environment or credential material to an external endpoint.

Public source snippet (untrusted):

```javascript
L7: `:case"\u2028":case"\u2029":return H(),"";case"\r":return H(),wr()===`
L8: `&&H(),"";case"1":case"2":case"3":case"4":case"5":case"6":case"7":case"8":case"9":throw Be(H());case void 0:throw Be(H())}return H()}function Sw(){let t="",e=wr();if(!Ge.isHexDigit...
L9: `+s;w=v.join(T),k=`{
...
L13: `+s+w+`,
L14: `+b+"]"}return o.pop(),s=b,v}}});var Ac=J((gO,np)=>{var Dw=ep(),vw=rp(),Rw={parse:Dw,stringify:vw};np.exports=Rw});var Ap=J((fF,Qw)=>{Qw.exports={"0":"\u2080","1":"\u2081","2":"\u2...
L15: `)||n,code:s,status:o},e.data.error)}return Object.assign({message:n,code:s,status:o},e.data.error)}}return{message:n,code:e.status,status:e.statusText}}};Dt.GaxiosError=Nc;functio...
L16: `).join(`
...
L49: `+_+"}":"{"+b.join(",")+"}",n=_,A}}typeof Tm.stringify!="function"&&(Tm.stringify=function(f
```

### 7. High: Command Output Exfiltration
- **Category:** Source
- **Confidence:** 82.0%
- **Path:** dist/esm/server.mjs
- **Public source:** [View source](<https://unpkg.com/@caeliq/llms@1.0.56/dist/esm/server.mjs>)

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

Public source snippet (untrusted):

```javascript
L49: `+_+"}":"{"+b.join(",")+"}",n=_,A}}typeof Tm.stringify!="function"&&(Tm.stringify=function(f,d,p){var y;if(n="",o="",typeof p=="number")for(y=0;y<p;y+=1)o+=" ";else typeof p=="stri...
L50: `,r:"\r",t:"	"},s,a=function(_){throw{name:"SyntaxError",message:_,at:r,text:s}},u=function(_){return _&&_!==n&&a("Expected '"+_+"' instead of '"+n+"'"),n=s.charAt(r),r+=1,n},c=fun...
L51: Supported algorithms are:
...
L60: ${A}
L61: `+await t.crypto.sha256DigestHex(y),_=await XT(t.crypto,t.securityCredentials.secretAccessKey,a,t.region,n),b=await ls(t.crypto,_,g),m=`${E_} Credential=${t.securityCredentials.acc...
L62: To learn more about authentication and Google APIs, visit:
```

### 8. High: Cloud Metadata Access
- **Category:** Source
- **Confidence:** 84.0%
- **Path:** dist/esm/server.mjs
- **Public source:** [View source](<https://unpkg.com/@caeliq/llms@1.0.56/dist/esm/server.mjs>)

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

Public source snippet (untrusted):

```javascript
L7: `:case"\u2028":case"\u2029":return H(),"";case"\r":return H(),wr()===`
L8: `&&H(),"";case"1":case"2":case"3":case"4":case"5":case"6":case"7":case"8":case"9":throw Be(H());case void 0:throw Be(H())}return H()}function Sw(){let t="",e=wr();if(!Ge.isHexDigit...
L9: `+s;w=v.join(T),k=`{
...
L13: `+s+w+`,
L14: `+b+"]"}return o.pop(),s=b,v}}});var Ac=J((gO,np)=>{var Dw=ep(),vw=rp(),Rw={parse:Dw,stringify:vw};np.exports=Rw});var Ap=J((fF,Qw)=>{Qw.exports={"0":"\u2080","1":"\u2081","2":"\u2...
L15: `)||n,code:s,status:o},e.data.error)}return Object.assign({message:n,code:s,status:o},e.data.error)}}return{message:n,code:e.status,status:e.statusText}}};Dt.GaxiosError=Nc;functio...
L16: `).join(`
...
L49: `+_+"}":"{"+b.join(",")+"}",n=_,A}}typeof Tm.stringify!="function"&&(Tm.stringify=function(f
```

### 9. Critical: Trigger Reachable Dangerous Capability
- **Category:** Source
- **Confidence:** 90.0%
- **Path:** dist/cjs/server.cjs
- **Public source:** [View source](<https://unpkg.com/@caeliq/llms@1.0.56/dist/cjs/server.cjs>)

A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.

Public source snippet (untrusted):

```javascript
Trigger-reachable chain: manifest.main -> dist/cjs/server.cjs
L7: `:case"\u2028":case"\u2029":return H(),"";case"\r":return H(),Er()===`
L8: `&&H(),"";case"1":case"2":case"3":case"4":case"5":case"6":case"7":case"8":case"9":throw Ne(H());case void 0:throw Ne(H())}return H()}function Ww(){let t="",e=Er();if(!Ge.isHexDigit...
L9: `+s;w=v.join(T),k=`{
...
L13: `+s+w+`,
L14: `+b+"]"}return o.pop(),s=b,v}}});var zc=J((W1,vp)=>{var Vw=Ap(),Kw=Dp(),Yw={parse:Vw,stringify:Kw};vp.exports=Yw});var Qp=J((IO,hE)=>{hE.exports={"0":"\u2080","1":"\u2081","2":"\u2...
L15: `)||n,code:s,status:o},e.data.error)}return Object.assign({message:n,code:s,status:o},e.data.error)}}return{message:n,code:e.status,status:e.statusText}}};Rt.GaxiosError=tl;functio...
L16: `).join(`
...
L49: `+_+"}":"{"+b.join(",")+"}",n=
```

### 10. Low: High Entropy Strings
- **Category:** Supply Chain
- **Confidence:** 55.0%

Package source contains high-entropy string patterns.

### 11. Low: Url Strings
- **Category:** Supply Chain
- **Confidence:** 65.0%

Package source contains URL literals.

### 12. Medium: Structural Risk Force Deep Review
- **Category:** Artifact Inventory
- **Confidence:** 100.0%

Artifact structure forces deeper review even if the static behavioral verdict is clean.

## Dependencies and install lifecycle
- **Lifecycle scripts present:** No

- **Dependencies:** 18
- **Optional dependencies:** 0
- **Peer dependencies:** 0
- **Development dependencies:** 4
- **Published dependency-graph edges:** 18

### Published dependency entries
- @anthropic-ai/sdk ^0.54.0 (Dependency)
- @caeliq/ccr-shared ^2.0.2 (Dependency)
- @cursor/sdk ^1.0.23 (Dependency)
- @fastify/cors ^11.0.1 (Dependency)
- @google/genai ^1.7.0 (Dependency)
- @huggingface/tokenizers ^0.0.6 (Dependency)
- dotenv ^16.5.0 (Dependency)
- fastify ^5.4.0 (Dependency)
- fastify-plugin ^5.1.0 (Dependency)
- google-auth-library ^10.1.0 (Dependency)
- json5 ^2.2.3 (Dependency)
- jsonrepair ^3.13.0 (Dependency)
- latex-to-unicode ^0.1.0 (Dependency)
- lru-cache ^11.2.2 (Dependency)
- openai ^5.6.0 (Dependency)
- tiktoken ^1.0.21 (Dependency)
- undici ^7.10.0 (Dependency)
- uuid ^11.1.0 (Dependency)

## Package metadata
- **Package:** @caeliq/llms
- **Ecosystem:** npm
- **Version:** 1.0.56
- **License:** MIT
- **Version published:** 2026-07-20T11:24:38.456Z
- **Package first seen:** 2026-07-20T11:26:56.810Z
- **Package last seen:** 2026-10-29T23:15:59.279Z
- **Known versions:** 15
- **Latest version:** 1.0.74
- **Appeal under review:** No
- **Description:** A universal LLM API transformation server
- **Author:** caeliq
- **Maintainers:** oakimov
- **Keywords:** llm, anthropic, openai, gemini, transformer, api
- **Artifact files:** 7
- **Artifact unpacked size:** 5,837,984 bytes
- **Artifact signatures:** 1
- **Attestations:** Yes
- **Provenance:** https://slsa.dev/provenance/v1

## References
- [HTML security report](<https://firewall.lpm.dev/npm/@caeliq/llms/v/1.0.56>)
- [Repository](<https://github.com/oakimov/claude-code-router>)
- [Homepage](<https://github.com/oakimov/claude-code-router#readme>)
- [Issues](<https://github.com/oakimov/claude-code-router/issues>)
