Back to skill

Security audit

基于AI的SBTI技能测试

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its SEO planning purpose, but its runner can execute shell code from a configurable secrets file before generating the plan.

Install only if you control the env file and CSV inputs. Run it as an unprivileged user, do not point AI_SBTI_SKILL_TEST_ENV_FILE at untrusted files, and review generated Markdown as untrusted content when prospect data comes from outside sources.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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 (2)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/run_weighted_backlink_plan.sh:5
Finding

Arbitrary Shell Code Execution Through Sourced Environment File

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/build_weighted_backlink_plan.py:202
Finding

Markdown Content Injection Through Unsanitized CSV Fields

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding

The skill clearly instructs the agent to read and write local files and run scripts that may perform network-related SEO operations, but it does not declare any explicit tool scope or allowed permissions in the metadata. This creates an authorization ambiguity where an execution environment may grant broader capabilities than reviewers expect, increasing the risk of unintended file access, secret handling, or outbound actions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The main skill heading is presented in Chinese ('基于AI的SBTI技能测试') while the rest of the file is otherwise English-oriented, and the document does not state that Chinese is optional or required for a justified region-specific use case. This can violate a language/locale policy when a skill implicitly constrains presentation or interaction language without user choice.

Content

No source excerpt is available for this finding.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 66)May include surrounding context.

mkdir -p /root/.openclaw/workspace/.secrets cp skills/ai-sbti-skill-test/config/ai-sbti-skill-test.env.example
/root/.openclaw/workspace/.secrets/ai-sbti-skill-test.env chmod 600 /root/.openclaw/workspace/.secrets/ai-sbti-skill-test.env

text

Generate weighted execution plan:

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The script automatically sources an environment file from a fixed secrets path before running an SEO-planning task, creating an unnecessary trust boundary and exposing the process to any variables or shell-side effects defined there. Even though the path is quoted and sourced conditionally, loading unrelated secrets into the runtime can leak sensitive data to child processes or let a tampered env file influence behavior unexpectedly.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The script silently sources a potentially sensitive environment file with source, which executes shell content rather than merely parsing key-value pairs. If that file is modified, replaced, or unexpectedly populated with shell commands, those commands run in the script's context, making this more dangerous than simple configuration loading—especially since an SEO skill has no obvious need to consume privileged secrets from a workspace secrets directory.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The script defines output paths for markdown and JSON artifacts and creates their parent directories, which indicates it will write files to the filesystem. There is no confirmation prompt, user-facing echo/log message, or explanatory comment/docstring in this file disclosing that it will create or overwrite these outputs.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The display name is in Chinese while the short description and default prompt are in English, and the file provides no indication that language is selectable by the user. This can create a language/locale policy issue because the skill implicitly imposes inconsistent language behavior without opt-in or documented justification.

Content

No source excerpt is available for this finding.