osborn@0.9.180
Voice AI coding assistant - local agent that connects to Osborn frontend
AI Security Review
scanned 1mo ago · by lpm-firewall-aiNo automatic npm install hook, self-dependency chain, credential exfiltration, or stealth payload was confirmed. The package nevertheless contains a user-triggered remote code installation path and import-time global Claude configuration changes.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
A user or connected frontend sends the start_ide command; importing the agent loads the Claude integration.
Impact
A compromised installer host could execute arbitrary commands; the agent changes Claude configuration outside its own package directory.
Mechanism
Remote shell installer plus global AI-agent configuration mutation.
AI rationale
No automatic npm install hook, self-dependency chain, credential exfiltration, or stealth payload was confirmed. The package nevertheless contains a user-triggered remote code installation path and import-time global Claude configuration changes.
Evidence
dist/index.jsdist/claude-llm.js~/.claude/settings.json
Network endpoints1
code-server.dev/install.sh
Decision evidence
public snapshotAI called this Suspicious at 91.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning
- An explicit IDE-start command can download an installer and pipe it directly to a shell.
- Importing the Claude integration changes the user's global Claude settings file.
- The remote installer is triggered from a frontend command, rather than during npm installation.
Evidence against
- The manifest has only a prepublish hook, so registry installation has no install lifecycle execution.
- The observed API calls use configured service credentials for the assistant's voice and meeting features; no secret-harvesting upload path was confirmed.
- Skill writes are scoped to a session directory and sanitize the supplied skill name.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
9 flagged · loading sourcebin/cli.jsView file
2L3: import { spawn } from 'child_process'
L4: import { fileURLToPath } from 'url'
High
dist/jsonl-search.jsView file
34// @vscode/ripgrep bundles the rg binary — this is our own dependency
L35: const { rgPath } = require('@vscode/ripgrep');
L36: if (rgPath && existsSync(rgPath))
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/jsonl-search.jsView on unpkg · L34dist/index.jsView file
2459const which = await new Promise((res) => {
L2460: const p = spawn('which', ['code-server']);
L2461: let out = '';
...
L2468: await new Promise((res, rej) => {
L2469: const install = spawn('sh', ['-c', 'curl -fsSL https://code-server.dev/install.sh | sh'], {
L2470: env: { ...process.env },
L2471: stdio: ['ignore', 'pipe', 'pipe'],
Findings
5 Critical3 High5 Medium5 Low
CriticalSame File Env Network Executiondist/index.js
CriticalCredential Exfiltrationdist/index.js
CriticalCommand Output Exfiltrationdist/index.js
CriticalDownload Executedist/index.js
CriticalPrevious Version Dangerous Deltadist/index.js
HighChild Processbin/cli.js
HighShell
HighSemantic Analysis Limiteddist/session-access.js#virtual:normalized:round1
MediumDynamic Requiredist/jsonl-search.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperdeploy.sh
MediumStructural Risk Force Deep Review
Affected versions and remediation
This report applies to osborn@0.9.180.
See all versions of osborn for other recorded verdicts.
- Review the evidence and your use of osborn@0.9.180 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.