Back to skill

Security audit

Openclaw User Profiler

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its profile-and-recommendation purpose, but it should be reviewed because it persists user profile data and encourages unpinned third-party skill installs.

Install only if you are comfortable with a skill that creates or updates user.md profile memory and recommends external skills. Before running any suggested npx skills add command, verify the source, version, publisher, and permissions of that third-party skill, and avoid storing secrets, IDs, health, financial, or other sensitive data in user.md.

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:136
Finding

Unpinned Third-Party Skill Installation Creates a Supply-Chain Risk

Content
View full analysis
` ``` After the user picks, provide concrete install instructions. ``` `references/role-skill-catalog.md:17`: ```markdown | 📦 External | Community Skill (GitHub / agentskills.sh) | `npx skills add ` | ``` `references/role-skill-catalog.md:751`: ```markdown **How to install**: `npx skills add ` ``` `references/role-skill-catalog.md:765-768`: ```markdown 1. **Add a new Skill**: Search with `npx skills find` → evaluate quality → add to the relevant role 2. **Deprecate a Skill**: Remove promptly or note a replacement 3. **Add a new role**: Append a new section at the end of the appropriate category, update inheritance relationships 4. **Update external Skills**: Run `npx skills check` periodically to check for updates ``` ### Technical Analysis The Skill recommends numerous community-maintained dependencies and generates `npx skills add` commands without pinning the `skills` CLI version, repository commit, immutable release, or integrity digest. The reviewed project does not vendor or audit the installed content. Consequently, the effective dependency content can change after this project has been reviewed. Risk sources include compromise of an npm package, repository, maintainer account, mutable branch, or similarly named package. A malicious dependency could introduce executable code or hostile Skill instructions into the user's Agent environment. The installation feature is relevant to the declared recommendation functionality, but unrestricted installation from mutable community sources exceeds the minimum trust necessary merely to recommend Skills. Recommendations can be provided without automatica ...[truncated 1737 chars]
Remediation
View remediation
add ``` 11. Re-audit pinned dependencies before updating them and document the reviewed revision in the catalog. ]]>
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (12)

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
85% confidence
Finding

Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Content

Scanner excerpt · SKILL.md (reported line 152)May include surrounding context.

md
1. **Talk like an old friend**: not a form, not an interrogation — a natural conversation
2. **Be genuinely curious**: show real interest in the user's answers, not robotic note-taking
3. **Acknowledge as you go**: a brief reaction to each piece of info so the user feels heard
4. **Don't judge**: you're learning who they are, not grading their choices

### Tone reference (vary each time — never copy verbatim)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger list includes broad natural-language phrases like 'know me', 'profile me', and role statements such as 'I'm an engineer', which can cause the skill to activate during ordinary conversation. Unintended invocation is especially sensitive here because the skill reads and writes user.md and may scan installed skills, so accidental activation can lead to unnecessary collection or modification of profile data.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill workflow instructs the agent to read existing user.md, write profile data, and scan the user's skills directory, but the user-facing description does not prominently warn about these behaviors up front. This weakens informed consent and increases the chance users trigger the skill without realizing it will access local profile files and enumerate installed tools.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The skill recommends installing packages via npx skills add <package> without constraining or pinning a specific package version. If a package is updated maliciously or a dependency is compromised, users may install unreviewed code at recommendation time, creating a supply-chain risk.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The catalog steers users toward a very large set of third-party skills, including ones with operational, deployment, automation, and external-service capabilities unrelated to simple profiling. In the context of a profiling/recommendation skill, this broad recommendation surface can become an indirect capability-escalation path by normalizing installation of powerful external skills without strong trust boundaries or scope limits.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The entry describes i18n-localization as 'Chinese-to-English localization in Reddit style,' which hard-codes a language direction in natural language. This can violate language/locale policy because it prescribes a specific language behavior without indicating user choice or a documented region-specific constraint.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file explicitly instructs users how to install third-party skills even though the parent skill's purpose is user profiling and role-based recommendation. That creates unnecessary pressure toward package installation and expands the attack surface beyond the justified scope of the feature.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The maintenance section instructs active discovery and update workflows (find, check) that go beyond static recommendation content. In this skill context, encouraging continual expansion and updating of external skills increases exposure to unreviewed or changed third-party packages without being necessary for profiling users.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.