pando-ai@1.2.4
pandō helps coding agents navigate code as a graph, make coordinated changes, and keep every committed step. It stores TypeScript and JavaScript projects locally as versioned programs, rendering ordinary files for existing tools. Automatic semantic commit
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis is a guarded agent-extension setup rather than confirmed malware. Its automatic tool-approval configuration creates a concrete agent-control risk.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
A user runs the package setup flow and accepts the launcher installation.
Impact
This expands an AI-agent control surface and can reduce normal tool-approval safeguards during launcher use.
Mechanism
The launcher configures an enabled, required MCP server for Codex and sets its tool approval mode to approve.
AI rationale
This is a guarded agent-extension setup rather than confirmed malware. Its automatic tool-approval configuration creates a concrete agent-control risk.
Evidence
dist/cli.js
Decision evidence
public snapshotAI called this Suspicious at 87.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning
- The package metadata has no npm install lifecycle hook.
- The CLI presents an opt-in prompt before installing launchers for Codex and Claude.
- The generated Codex configuration enables a required MCP server and sets tool approval to approve.
Evidence against
- No automatic npm lifecycle execution was found.
- The launcher setup is user-invoked and includes an interactive opt-in prompt.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsEvalFilesystemNativeBindingsNetwork
HighEntropyStringsMinifiedUrlStrings
NoLicense
Oversized source lightweight scan
Source & flagged code
19 flaggeddist/typescript-reference-analyzer.cjsView file
1"use strict";var Ke=Object.create;var le=Object.defineProperty;var He=Object.getOwnPropertyDescriptor;var Ze=Object.getOwnPropertyNames;var Qe=Object.getPrototypeOf,et=Object.proto...
L2: `)}catch{}try{ct(),oe.default.appendFileSync(Oe(),a+`
High
Child Process
Package source references child process execution.
dist/typescript-reference-analyzer.cjsView on unpkg · L1dist/workers/typescript-operational-producer.mjsView file
3245${u}`)}}));let r=await D.loadWebAssemblyModule(await n,{loadAsync:!0}),o=Object.keys(r),s=o.find(c=>VR.test(c)&&!c.includes("external_scanner_"));if(!s)throw console.log(`Couldn't ...
L3246: ${JSON.stringify(o,null,2)}`),new Error("Language.load failed: no language function found in WASM file");let a=r[s]();return new UI(ki,a)}},KR=(()=>{var _scriptName=import.meta.url...
L3247: project.source_file_affecting_options_key,membership.compiler_options_json
High
Eval
Package source references dynamic code evaluation.
dist/workers/typescript-operational-producer.mjsView on unpkg · L3245Findings
3 Critical10 High8 Medium6 Low
CriticalSame File Env Network Executiondist/cli.js
CriticalCredential Exfiltrationdist/cli.js
CriticalTrigger Reachable Dangerous Capabilitydist/cli.js
HighChild Processdist/typescript-reference-analyzer.cjs
HighEvaldist/workers/typescript-operational-producer.mjs
HighCredential Redirect Persistencedist/cli.js
HighCommand Output Exfiltrationdist/workers/typescript-operational-producer.mjs
HighObfuscated Payload Loaderdist/workers/typescript-operational-producer.mjs
HighCross File Remote Execution Contextdist/workers/native-authority-worker.mjs
HighTrigger Reachable Credential Exfiltrationdist/cli.js
Affected versions and remediation
This report applies to pando-ai@1.2.4.
See all versions of pando-ai for other recorded verdicts.
- Review the evidence and your use of pando-ai@1.2.4 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.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.