registry / @awareness.market/local / 0.13.0
@awareness.market/local@0.13.0
Local-first AI agent memory system. No account needed.
AI Security Review
scanned 22d ago · by lpm-firewall-aiThe static warning is not borne out by the inspected source: the postinstall fallback is limited to the declared Transformers dependency, and the CLI and browser process calls have package-aligned purposes and constraints. The agent guidance appears in project-configuration metadata; no confirmed automatic agent configuration write was identified.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
Package installation, explicit daemon startup, or a request to open an Awareness Market URL.
Impact
No confirmed unauthorized data transfer, agent configuration mutation, or other attack impact was established.
Mechanism
The install helper checks for Transformers and installs it only if missing. The CLI relaunches its own daemon, while the browser handler validates the destination host.
AI rationale
The static warning is not borne out by the inspected source: the postinstall fallback is limited to the declared Transformers dependency, and the CLI and browser process calls have package-aligned purposes and constraints. The agent guidance appears in project-configuration metadata; no confirmed automatic agent configuration write was identified.
Evidence
package.jsoninstall-transformers.mjsawareness-spec.jsonsrc/spec/awareness-spec.jsonbin/awareness-local.mjssrc/daemon/api-handlers.mjs
Decision evidence
public snapshotAI called this Clean at 91.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The postinstall hook runs a helper that exits when the declared Transformers dependency is available; its fallback installs that dependency.
- Both spec files describe project-level Claude Code rules and MCP configuration, with guidance for using this package’s memory features.
- The CLI child process relaunches the same daemon command with the selected project and port.
- The browser-opening handler restricts its URL to the package’s Awareness Market host before invoking the platform browser opener.
Evidence against
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
9 flaggedpackage.jsonView file
•scripts.postinstall = node install-transformers.mjs
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkgbin/awareness-local.mjsView file
16L17: import { spawn } from 'node:child_process';
L18: import fs from 'node:fs';
High
Child Process
Package source references child process execution.
bin/awareness-local.mjsView on unpkg · L16bin/cli.cjsView file
3// Wrapper to load the ESM entry point from CommonJS-compatible bin
L4: const { pathToFileURL } = require('node:url');
L5: const { join } = require('node:path');
Findings
1 Critical8 High4 Medium5 Low
CriticalPrevious Version Dangerous Deltasrc/daemon/api-handlers.mjs
HighInstall Time Lifecycle Scriptspackage.json
HighChild Processbin/awareness-local.mjs
HighShell
HighSame File Env Network Executionbin/awareness-local.mjs#virtual:normalized:round1
HighCommand Output Exfiltrationbin/awareness-local.mjs#virtual:normalized:round1
HighSandbox Evasion Gated Capabilitysrc/daemon/api-handlers.mjs
HighTrigger Reachable Command Output Exfiltrationbin/awareness-local.mjs#virtual:normalized:round1
HighSemantic Analysis Limitedbin/awareness-local.mjs#virtual:normalized:round1
MediumDynamic Requirebin/cli.cjs
MediumNetwork
Affected versions and remediation
This report applies to @awareness.market/local@0.13.0.
See all versions of @awareness.market/local for other recorded verdicts.
- Review the evidence and your use of @awareness.market/local@0.13.0 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.