Back to skill

Security audit

Skills

Security checks for vulnerabilities and agentic risk

Overview

This skill is purpose-aligned but should be reviewed because it lets an external analysis service shape current and future agent behavior without clear user approval controls.

Install only if you are comfortable with ClawTrace analyzing your agent trajectories and with the agent using those recommendations. Before use, require explicit approval for each recommendation, do not allow changes to safety rules or unrelated goals, redact sensitive trace content where possible, and review any MEMORY.md entry before it is written.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (4)

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:109
Finding

Remote Recommendations Can Hijack the Current Agent Session

Content
View full analysis
Remediation
View remediation

T01 · Skill Instruction Hijacking

Error
Location
self-evolve.md:109
Finding

Remote Recommendations Can Hijack the Current Agent Session in the Mirrored Skill File

Content
View full analysis
Remediation
View remediation

T02 · Agent Memory Poisoning

Error
Location
SKILL.md:115
Finding

Untrusted Remote Recommendations Are Written to Persistent Agent Memory

Content
View full analysis
Remediation
View remediation

T02 · Agent Memory Poisoning

Error
Location
self-evolve.md:115
Finding

Untrusted Remote Recommendations Are Written to Persistent Memory by the Mirrored Skill File

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (12)

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

In context, the skill does more than use a name: it instructs the agent to analyze its own trajectories, apply changes to the current session immediately, and persist learnings for future behavior. That creates a real self-modification pathway where an external service can shape future agent conduct, increasing risk if the service is compromised, incorrect, or manipulated by poisoned trajectory data.

Content

Scanner excerpt · SKILL.md (reported line 8)May include surrounding context.

md
metadata: {"openclaw":{"emoji":"🔬","os":["linux","darwin","win32"],"homepage":"https://clawtrace.ai","requires":{"env":["CLAWTRACE_OBSERVE_KEY"]}}}
---

# ClawTrace Self-Evolve Skill

You have access to ClawTrace, an observability platform that captures and analyzes all your trajectories. Use this skill to understand your own performance and improve over time.

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · SKILL.md (reported line 2)May include surrounding context.

md
---
name: ClawTrace Self-Evolve
description: Ask Tracy to analyze your recent trajectories and improve your agent behavior based on data-driven recommendations.
version: 1.0.0
metadata: {"openclaw":{"emoji":"🔬","os":["linux","darwin","win32"],"homepage":"https://clawtrace.ai","requires":{"env":["CLAWTRACE_OBSERVE_KEY"]}}}

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · self-evolve.md (reported line 2)May include surrounding context.

md
---
name: ClawTrace Self-Evolve
description: Ask Tracy to analyze your recent trajectories and improve your agent behavior based on data-driven recommendations.
version: 1.0.0
metadata: {"openclaw":{"emoji":"🔬","os":["linux","darwin","win32"],"homepage":"https://clawtrace.ai","requires":{"env":["CLAWTRACE_OBSERVE_KEY"]}}}

Self-Modification

High
Category
Rogue Agent
Confidence
90% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · self-evolve.md (reported line 8)May include surrounding context.

md
---
name: ClawTrace Self-Evolve
description: Ask Tracy to analyze your recent trajectories and improve your agent behavior based on data-driven recommendations.
version: 1.0.0
metadata: {"openclaw":{"emoji":"🔬","os":["linux","darwin","win32"],"homepage":"https://clawtrace.ai","requires":{"env":["CLAWTRACE_OBSERVE_KEY"]}}}

External Transmission

Medium
Category
Data Exfiltration
Confidence
98% confidence
Finding

This skill explicitly sends data to an external endpoint, which is a genuine external transmission path. Even if expected functionality, it is security-relevant because the payload can include trace and session identifiers and queries about prior runs, potentially exposing sensitive operational data to a third party.

Content

Scanner excerpt · SKILL.md (reported line 23)May include surrounding context.

md
## How to Call the Endpoint

Send a POST request to `https://api.clawtrace.ai/v1/evolve/ask`.

Authentication uses your observe key, which is already set in your environment as `CLAWTRACE_OBSERVE_KEY`.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill directs the agent to send trajectory data to an external service but does not clearly warn that prompts, outputs, failures, and other run metadata may leave the local environment. Because trajectory data can contain sensitive user content, this creates a real privacy and data-governance risk even if the service is legitimate.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
99% confidence
Finding

The embedded code performs an authenticated POST to a third-party API using an environment-sourced bearer token, confirming real outbound data flow. This is dangerous in contexts where traces may contain secrets, regulated data, or proprietary prompts, because the skill normalizes transmitting them off-box without inline safeguards.

Content

Scanner excerpt · SKILL.md (reported line 47)May include surrounding context.

md
with httpx.stream(
        "POST",
        "https://api.clawtrace.ai/v1/evolve/ask",
        headers={
            "Authorization": f"Bearer {observe_key}",
            "Content-Type": "application/json",

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

This finding reflects an explicit instruction to transmit data to an external domain, creating a data exfiltration path outside the local trust boundary. In context, the endpoint is the intended service, but the danger is that users may not realize that operational traces and identifiers are being sent off-platform for analysis.

Content

Scanner excerpt · self-evolve.md (reported line 23)May include surrounding context.

md
## How to Call the Endpoint

Send a POST request to `https://api.clawtrace.ai/v1/evolve/ask`.

Authentication uses your observe key, which is already set in your environment as `CLAWTRACE_OBSERVE_KEY`.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill instructs the agent to send questions plus trace or session identifiers to a third-party API that analyzes recent trajectories, but it does not provide a clear user-facing warning or consent step about external data sharing. Because trajectory data may contain prompts, tool inputs, outputs, and other sensitive operational context, this can result in unreviewed disclosure of sensitive information.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
94% confidence
Finding

The embedded code performs a live streamed POST request to a third-party API using a bearer token from the environment, which operationalizes external transmission rather than merely describing it. If used without guardrails, this can leak sensitive trace context or normalize sending internal observability data to an external service without sufficient review.

Content

Scanner excerpt · self-evolve.md (reported line 47)May include surrounding context.

md
with httpx.stream(
        "POST",
        "https://api.clawtrace.ai/v1/evolve/ask",
        headers={
            "Authorization": f"Bearer {observe_key}",
            "Content-Type": "application/json",

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill goes beyond passive analysis and directs the agent to modify local state, including trimming session history and writing to MEMORY.md. That creates a trust boundary issue: an external service's recommendations can directly influence persistent memory and runtime behavior, which could cause unsafe or unintended changes if the advice is wrong, manipulated, or over-broad.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs immediate writes to MEMORY.md without warning the user that a local file will be modified. Silent persistence of externally influenced recommendations can create integrity and auditability issues, especially if later sessions trust MEMORY.md as authoritative context.

Content

No source excerpt is available for this finding.