Back to skill

Security audit

Safe Driving Behavior Analyzer | 安全驾驶行为分析工具

Security checks for vulnerabilities and agentic risk

Overview

The scanners report several privacy, account-handling, and behavior-scope concerns in this driver-video analysis skill, so ClawScan classifies it as suspicious; the evidence does not establish malicious intent.

Consider the suspicious verdict and the reported risks before installing. In particular, decide whether you are comfortable with video being processed remotely, account registration and credential storage, and a no-analysis option that may not prevent analysis. The scan was partial because a referenced local path was missing, so it does not provide complete evidence.

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 (1)

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/drive_analysis.py
Finding

Sender identity is discarded in favor of shared cached credentials

Content
View full analysis

Impact

Another sender can receive the fallback account's driving-analysis history and report links. Analysis requests under the same conditions can also be attributed to the shared account.

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
Findings (23)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding

The supplied chunk implements user-account persistence rather than video processing, unsafe-driving detection, or report generation. Generic database access could be a supporting implementation detail, but this code specifically manages user identities, demographic fields, email addresses, and token fields, with no demonstrated connection to driving analysis. These account-management capabilities are not represented in the description. The chunk does not show external transmission or establish that sensitive values are actually populated.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

Ordinary file, JSON, polling, and HTTP helpers could support the declared video-analysis workflow and are not themselves mismatches. However, this chunk also implements automatic account enrollment, persistent identity/token management, and unrestricted destination handling for credential-bearing requests—capabilities not represented by the description. No video processing, unsafe-driving detection, or report generation appears in this chunk, although those functions could exist elsewhere. The finding concerns these additional capabilities, not merely the absence of analysis logic in a shared utility module.

Content

No source excerpt is available for this finding.

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · SKILL.md (reported line 3)May include surrounding context.

md
---
name: "drive-analysis"
description: "Analyzes videos of vehicle drivers to identify unsafe driving behaviors. It generates professional analysis reports to help enhance road safety awareness. | 安全驾驶行为分析工具,针对机动车驾驶人员的驾驶行为进行视频分析,识别不安全驾驶行为,输出专业分析报告,提升道路交通安全意识"
version: "1.0.27"
license: "MIT-0"
---

# 🚗 Safe Driving Behavior Analyzer | 安全驾驶行为分析工具
> **智能分析中枢** · 图片/视频智能分析 · 结构化报告 · 历史报告云端查询

---

## 🧭 技能概览 | Ove

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, file_write, network, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The analysis workflow lacks a privacy warning about remote processing of driving footage.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

Although the documentation mentions APIs and cloud reports, it does not clearly warn that local footage is transferred to a remote processor or explain retention and access implications. Driving footage can reveal faces, passengers, locations, and private conversations, making automatic processing without informed authorization a meaningful privacy risk.

Content

No source excerpt is available for this finding.

The documented no-analysis mode still executes video analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The parser accepts --export-env-only and promises not to perform analysis, but execution never checks the flag before identity initialization or analysis. When an input is supplied, a user relying on this mode can unintentionally submit sensitive footage for remote processing.

Content

No source excerpt is available for this finding.

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · skills/smyx_analysis/scripts/skill.py (reported line 28)May include surrounding context.

python
result_json = JsonUtil.parse(result_json_pure_text, result_json_pure_text)

        result_json_common_ai_response = result_json.get("commonAiResponse") if isinstance(result_json,
                                                                                           dict) else result_json
        if result_json_common_ai_response:
            result_json = result_json_common_ai_response

Whitespace Padding

Medium
Category
Prompt Injection
Confidence
70% confidence
Finding

Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Content

Scanner excerpt · skills/smyx_analysis/scripts/skill.py (reported line 33)May include surrounding context.

python
result_json = result_json_common_ai_response

        result_json_health_ai_response = result_json.get("healthAiResponse") if isinstance(result_json,
                                                                                           dict) else result_json
        if result_json_health_ai_response:
            result_json = result_json_health_ai_response

Local video contents are submitted to an analysis API without disclosure of the upload.

Medium
Category
Output Handling
Confidence
92% confidence
Finding

The local-file branch reads the complete video and passes its contents as a file attachment to the analysis service without a visible upload notice or authorization check. Remote analysis is relevant to the skill, but relevance does not establish informed permission to disclose identifiable driving footage.

Content

No source excerpt is available for this finding.

The documented export-only option does not prevent video analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The export-only flag is parsed but never consulted, so supplying a video still reaches analyze_video despite the explicit no-analysis promise. This defeats a user's attempt to avoid remote processing and can cause unintended disclosure of sensitive video.

Content

No source excerpt is available for this finding.

The skill includes persistent user-account and credential management unrelated to driver-video analysis.

Medium
Category
Excessive Agency
Confidence
88% confidence
Finding

Account persistence can legitimately support cloud report history, but this implementation stores authentication tokens as ordinary strings alongside personal information in a workspace-shared SQLite database. No encryption or restrictive file-permission enforcement is visible, so another principal able to read that database could recover reusable credentials and potentially access sensitive reports.

Content

No source excerpt is available for this finding.

ai_chat documents an agent invocation and analysis result, but never invokes an agent.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The docstring states that this method calls openclaw agent through subprocess and returns an analysis result or session ID. However, both command construction and execution are commented out; the method instead accesses stderr on an empty dictionary, catches the resulting exception, and implicitly returns None. This contradicts the documented analysis behavior rather than merely omitting an implementation detail.

Content

No source excerpt is available for this finding.

FileUtil.open silently truncates existing files without disclosing its destructive behavior.

Medium
Category
Output Handling
Confidence
97% confidence
Finding

Opening a path with mode 'w' immediately truncates any existing file. This generically named helper has no confirmation, action log, docstring, or comment explaining that calling it can erase existing contents; exception tracing does not disclose successful truncation.

Content

No source excerpt is available for this finding.

Analysis requests can silently register an external account and persist its credentials.

Medium
Category
Excessive Agency
Confidence
90% confidence
Finding

An ordinary request can trigger a remote login or registration with silent and register enabled, sending the current identifier and persisting returned authentication tokens through UserDao. This introduces an undisclosed account-creation and credential-storage side effect into a driving-analysis workflow; the supplied source shows no explicit enrollment approval, although database protections and any upstream consent are unknown.

Content

No source excerpt is available for this finding.

HTTP requests can silently register a remote account and transmit the current user identifier.

Medium
Category
Output Handling
Confidence
92% confidence
Finding

The helper submits the selected username as both openId and mobile while enabling silent registration, so a request can disclose an existing user identifier and create a remote account without a distinct enrollment action. This is the same underlying consent issue as the other registration finding, not evidence of a separate exfiltration channel.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
85% confidence
Finding

This POST is the transmission point for the silent-registration flow: its payload includes the selected user identifier as openId and mobile. Network authentication is not inherently unsafe, but the automatic enrollment without a visible consent gate creates the privacy issue here; this call does not show driver-video transmission, and the destination configuration is not supplied.

Content

Scanner excerpt · skills/smyx_common/scripts/util.py (reported line 561)May include surrounding context.

python
"source": ConstantEnum.DEFAULT__SKILL_HUB_NAME
            }
            try:
                _response = requests.post(_url, json=_data)
                if _response.status_code == 200:
                    _response_json = _response.json()
                    if _response_json and _response_json.get("success"):

Payment failures redirect users into installing a separate payment skill.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

On HTTP 402, the code returns instructions to install 'smyx-payment' and initiate account recharge rather than simply reporting that analysis is unavailable. This introduces an undisclosed payment and additional-skill onboarding workflow beyond the manifest's video-analysis and safety-reporting purpose; the code recommends installation but does not itself perform it.

Content

No source excerpt is available for this finding.

Configuration loading silently creates directories and writes missing configuration files, including during module import.

Low
Category
Output Handling
Confidence
92% confidence
Finding

YamlUtil.load creates parent directories and writes a YAML file when the requested path is missing, without a prompt, log, docstring, or explanatory comment disclosing this side effect. BaseEnum.init_subclass invokes this loader during class creation (L044–L058), so importing this module can modify the filesystem without an explicit save request.

Content

No source excerpt is available for this finding.

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/dao.py (reported line 243)May include surrounding context.

python
username = getattr(model, "username", None)
            if username:
                column_names = self.__model__.__table__.columns.keys()
                update_data = {key: getattr(model, key) for key in column_names if key != "username"}
                return self.update_by_username(username, **update_data)

            return None

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/dao.py (reported line 353)May include surrounding context.

python
column_names = self.__model__.__table__.columns.keys()

            for key in column_names:
                value = getattr(model, key)
                setattr(instance, key, value)

            session.commit()

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/dao.py (reported line 326)May include surrounding context.

python
if filters:
                for key, value in filters.items():
                    query = query.filter(getattr(self.__model__, key) == value)

            if offset:
                query = query.offset(offset)

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/dao.py (reported line 441)May include surrounding context.

python
if filters:
                for key, value in filters.items():
                    query = query.filter(getattr(self.__model__, key) == value)

            return query.scalar()
        finally:

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · skills/smyx_common/scripts/util.py (reported line 30)May include surrounding context.

python
elif isinstance(cur, dict):
                        cur = cur.get(part, default)
                    else:
                        cur = getattr(cur, part)
                return cur
            except Exception:
                return default