registry / synapse-hub / 2.2.1
synapse-hub@2.2.1
Phien ban cu da duoc thu hoi. Vui long nang cap len synapse-hub@3.0.1 de su dung day du 51 skills va cac ban va bao mat moi nhat.
Enterprise AI Agent Intelligence Hub & Autonomous Skill Network
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe flagged paths contain static security-detection patterns and a local report-triage helper, not credential theft or a hidden payload. The CLI's routine update check fetches public version metadata and caches it locally.
Static reason
One or more suspicious static signals were detected.
Trigger
The version check runs on most CLI commands when its cache is stale; applying an update requires the explicit update command.
Impact
No harmful impact identified. The default check contacts update endpoints and stores version-check metadata locally.
Mechanism
The check requests public version metadata from the npm registry and GitHub, then may write a local update cache.
AI rationale
The flagged paths contain static security-detection patterns and a local report-triage helper, not credential theft or a hidden payload. The CLI's routine update check fetches public version metadata and caches it locally.
Evidence
~/.agent_memory/update_cache.json
Network endpoints2
registry.npmjs.orgraw.githubusercontent.com
Decision evidence
public snapshotAI called this Clean at 93.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The npm entry point dispatches the supplied arguments to the package's local Python CLI.
- The flagged credential and download patterns are static detection rules in the Skillspector scanner.
- The flagged shell helper searches local JUnit reports and screenshot filenames.
- The CLI's update check requests public package version metadata and uses a 24-hour cache.
Evidence against
- No install lifecycle scripts are declared in package.json.
- The inspected credential pattern is a regex matching AWS credential paths, not code that reads or sends credentials.
- The default update check sends no user or project data in the inspected request.
Behavioral surface
ChildProcessEnvironmentVarsFilesystemShell
UrlStrings
Source & flagged code
7 flaggedbin/agent-qa.jsView file
3// SPDX-License-Identifier: MIT
L4: // Project: Synapse-Hub (https://github.com/TeeVee998/Synapse-Hub)
L5:
...
L12:
L13: const { spawn, execSync } = require('child_process');
L14: const path = require('path');
...
L16:
L17: const REPO_ROOT = path.resolve(__dirname, '..');
L18: const CLI_PATH = path.join(REPO_ROOT, 'tools', 'cli.py');
...
L31: function resolvePython() {
L32: const candidates = process.platform === 'win32'
L33: ? ['python', 'py', 'python3']
High
Sandbox Evasion Gated Capability
Source gates dangerous network, credential, or execution behavior behind CI, host, platform, time, or geo fingerprint checks.
bin/agent-qa.jsView on unpkg · L369const args = [CLI_PATH, ...process.argv.slice(2)];
// Spawn child Python process inheriting standard IO
const child = spawn(python.command, args, {
stdio: 'inh
Low
Ai Review Evidence
The npm entry point dispatches the supplied arguments to the package's local Python CLI.
bin/agent-qa.jsView on unpkg · L69tools/tms_sync_bridge.pyView file
Findings
2 High3 Medium7 Low
HighSandbox Evasion Gated Capabilitybin/agent-qa.js
HighPayload In Excluded Dir.agents/skills/failure-autopsy-and-healing/scripts/triage_failures.sh
MediumEnvironment Vars
MediumShips Build Helpertools/tms_sync_bridge.py
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowUrl Strings
LowAi Review Evidencebin/agent-qa.js
LowAi Review Evidencetools/skillspector.py
LowAi Review Evidencetools/skillspector.py
LowAi Review Evidence.agents/skills/failure-autopsy-and-healing/scripts/triage_failures.sh
Affected versions and remediation
This report applies to synapse-hub@2.2.1.
See all versions of synapse-hub for other recorded verdicts.
- Review the evidence and your use of synapse-hub@2.2.1 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.