osborn@0.9.235
Voice AI coding assistant - local agent that connects to Osborn frontend
AI Security Review
scanned 26d ago · by lpm-firewall-aiThe automatic persistent Claude configuration mutation merits a warning under the explicit-user-command agent configuration policy. It is not supported as malware because the package lacks an install lifecycle trigger and no concrete exfiltration or payload execution was found.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
A user runs the package command that loads the main runtime.
Impact
The package can alter persistent Claude behavior for the local user without a separate configuration prompt.
Mechanism
Module-load code reads, creates if necessary, and rewrites the user's Claude settings file to set compaction and cleanup preferences.
AI rationale
The automatic persistent Claude configuration mutation merits a warning under the explicit-user-command agent configuration policy. It is not supported as malware because the package lacks an install lifecycle trigger and no concrete exfiltration or payload execution was found.
Evidence
package.jsondist/index.jsdist/claude-llm.js~/.claude/settings.json
Decision evidence
public snapshotAI called this Suspicious at 90.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning
- The package exposes command-line binaries and declares no install lifecycle hook.
- The main runtime imports the Claude LLM module.
- That module reads and modifies the user's Claude settings file, including compaction and retention values, and invokes this setup at module load.
- The settings change is triggered by running the package rather than during npm installation.
Evidence against
- No install, preinstall, or postinstall hook was declared.
- The inspected configuration change is limited to Claude settings and does not establish credential export or remote payload execution.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
8 flagged · loading sourcebin/recall.jsView file
5L6: import { spawn } from 'child_process'
L7: 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
2563const which = await new Promise((res) => {
L2564: const p = spawn('which', ['code-server']);
L2565: let out = '';
...
L2572: await new Promise((res, rej) => {
L2573: const install = spawn('sh', ['-c', 'curl -fsSL https://code-server.dev/install.sh | sh'], {
L2574: env: { ...process.env },
L2575: stdio: ['ignore', 'pipe', 'pipe'],
Findings
4 Critical3 High5 Medium5 Low
CriticalSame File Env Network Executiondist/index.js
CriticalCredential Exfiltrationdist/index.js
CriticalCommand Output Exfiltrationdist/index.js
CriticalDownload Executedist/index.js
HighChild Processbin/recall.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
LowNon Install Lifecycle Scripts
LowScripts Present
Affected versions and remediation
This report applies to osborn@0.9.235.
See all versions of osborn for other recorded verdicts.
- Review the evidence and your use of osborn@0.9.235 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.