osborn@0.9.178
Voice AI coding assistant - local agent that connects to Osborn frontend
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis is a dangerous runtime capability with unguarded global Claude configuration writes, but not a demonstrated malicious install-time or exfiltration chain. The scanner's claimed credential-exfiltration finding was not confirmed by source inspection.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
Run osborn; the installer additionally requires a start_ide request.
Impact
Claude configuration and skills can persist outside the package; the IDE feature trusts a remote installation script.
Mechanism
Runtime Claude configuration writes and a shell-piped remote installer.
AI rationale
This is a dangerous runtime capability with unguarded global Claude configuration writes, but not a demonstrated malicious install-time or exfiltration chain. The scanner's claimed credential-exfiltration finding was not confirmed by source inspection.
Evidence
dist/claude-llm.jsdist/index.js~/.claude/settings.json~/.claude/skills/learned-behaviors/SKILL.md
Network endpoints1
code-server.dev/install.sh
Decision evidence
public snapshotAI called this Suspicious at 88.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- Running the package changes global Claude settings at module load.
- A post-compaction hook writes model-derived content into global Claude skill files.
- An explicit IDE request downloads and runs a remote shell installer.
Evidence against
- The manifest has only a prepublish build hook, not an install lifecycle hook.
- The remote installer is reached only from the explicit start_ide handler.
- No source evidence shows environment or credential values being sent to an unrelated endpoint.
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@0.9.178.
See all versions of osborn for other recorded verdicts.
- Review the evidence and your use of osborn@0.9.178 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.