Back to skill

Security audit

ActivityClaw Plugin Usage

Security checks for vulnerabilities and agentic risk

Overview

The skill is coherent but asks users to install an unpinned third-party activity-tracking plugin that records broad agent activity into a local database and dashboard.

Review the ActivityClaw package and pin a known version before installing. Use it only if you are comfortable storing agent file activity, commands, web actions, messages, and sub-agent history locally, and confirm how to restrict dashboard access and delete the SQLite database when no longer needed.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:15
Finding
Unpinned Third-Party Package Installation## Vulnerability Details **File Location**: `SKILL.md`, lines 15–20 **Vulnerability Type**: Unpinned and unverifiable third-party dependency **Risk Level**: Medium ```markdown ## Prerequisites The ActivityClaw plugin must be installed: ```bash npm install -g @rmruss2022/activityclaw openclaw plugins install @rmruss2022/activityclaw ``` ``` ### Technical Analysis The skill directs users to install `@rmruss2022/activityclaw` globally and as an OpenClaw plugin without specifying an immutable version or integrity digest. The project contains no lockfile, vendored source, checksum, or plugin implementation that would allow the installed artifact to be verified against the version reviewed during this audit. As a result, the effective package payload can change after the skill has been reviewed. If the package publisher account or distribution channel is compromised, a malicious release could be delivered under the same package name. npm lifecycle scripts may execute during installation with the privileges of the invoking user, while a loaded OpenClaw plugin may operate within the host application's trust boundary. The package name and repository reference are consistent, and no typographical dependency confusion or known malicious payload is demonstrated by the available file. The risk arises from mutable, unpinned supply-chain installation rather than confirmed malicious intent. ### Attack Path 1. An attacker compromises the package publisher account, npm package, or associated release process. 2. The attacker publishes a malicious release under the existing `@rmruss2022/activityclaw` package name. 3. A user follows the documented commands without pinning a previously reviewed version. 4. npm retrieves the attacker-controlled current release. 5. Malicious lifecycle code may execute during global installation, or attacker-controlled plugin logic may execute when OpenClaw loads the plugin. 6. The malicious package ab ...[truncated 854 chars]
Remediation
## Remediation Suggestions 1. Pin the dependency to a specific, reviewed version rather than installing the latest mutable release. 2. Verify the downloaded package with an approved integrity digest or signed provenance before installation. 3. Record the exact dependency version and integrity value in a lockfile or equivalent reproducible installation manifest. 4. Prefer project-local installation over global installation where supported to reduce the affected scope. 5. Review the package contents, lifecycle scripts, transitive dependencies, and OpenClaw plugin entry points before enabling it. 6. Disable npm lifecycle scripts during acquisition when compatible, then explicitly run only reviewed setup operations. 7. Execute the plugin under a least-privileged account and restrict access to sensitive files, credentials, network destinations, and message data. 8. Establish a controlled update process in which each new package version is reviewed and approved before deployment.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill promotes use of an activity-tracking plugin and dashboard but does not warn users that it records sensitive operational telemetry such as file activity, commands, web actions, messages, and sub-agent behavior. This omission can lead users to enable or consult the feature without understanding the privacy and security implications of exposing local agent history, especially if the dashboard is accessible from the host or other local users.

Static analysis

No suspicious patterns detected.