Back to skill

Security audit

Plant Vitality Index | 植物整体活力指数(综合评分)

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because scanners report account and credential handling, media uploads, and behavior that does not clearly match its advertised plant-vitality analysis; the evidence is incomplete.

The verdict is suspicious, with high overall confidence. Before installing, consider whether you are comfortable with the reported account and credential handling, local persistence, and sending selected media to an external service without clear disclosure. The scanner completed analysis of the listed components, but marked coverage partial because a referenced local path was missing; behavior inside unavailable dependencies also cannot be confirmed.

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
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 (26)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied chunk implements user-account persistence rather than plant analysis. Its sys_user model includes email, birthday, sex, age, token, and open_token fields, and its DAO supports reading and modifying those records. No plant-image processing, environmental-data fusion, vitality scoring, or trend computation appears in this chunk. Database persistence could support a larger plant-monitoring application, but the account and token-management functionality shown is not established as a supporting requirement by the declared purpose. These operations occur when the DAO is instantiated or its methods are called; the chunk does not demonstrate network transmission or automatic execution of those operations.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The visible code is organized around MP4 video analysis and history listing. It does not collect or explicitly process plant environmental readings, chlorophyll indices, morphology, bud counts, or leaf-area growth, nor does it construct or validate a vitality score and trend. Its analysis behavior is delegated to skill.get_output_analysis(), whose implementation is not supplied, so plant-specific processing inside that dependency cannot be ruled out. Nevertheless, the supplied interface has a materially different apparent purpose and exposes an undeclared history-retrieval capability. Optional result-file writing is a supporting implementation detail, not itself a mismatch.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The declared plant-assessment behavior is not represented in the supplied implementation. No active external agent command is executed, and there is no visible plant-analysis logic or scoring output. Imported base classes and services could provide additional behavior, but their implementations are not supplied, so they cannot establish that the declared purpose is fulfilled. No malicious or unrelated external action is demonstrated by this chunk.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

Generic formatting, polling, and HTTP helpers could reasonably support a plant-monitoring skill and are not themselves mismatches. However, this chunk also implements an undisclosed account-registration, credential-management, and billing workflow, accessing local identity resources and remote authentication services beyond the stated evaluation purpose. No plant-feature fusion, vitality scoring, or trend calculation is present in the supplied chunk, although those functions could exist elsewhere. The mismatch is based on the concrete additional account and payment behavior, not merely the absence of scoring logic in a utility file.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

The supplied code explicitly requests a local video path or network video URL and forwards that input to an analysis service. Its response handling extracts generic commonAiResponse or healthAiResponse data without enforcing plant-specific scores or trends. Uploading inputs, polling, and exporting reports could be supporting implementation details, but the visible video-analysis interface is materially different from the declared multimodal plant-vitality evaluation. The imported API implementation is not supplied, so plant-specific behavior inside that service cannot be ruled out; it is not established by this chunk.

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: "smyx-plant-vitality-index-analysis"
description: "Using a plant-monitoring platform that periodically (e.g., daily) collects plant images, environmental data, and growth metrics (new bud count, leaf-area change, leaf color), an AI evaluation model fuses leaf color (chlorophyll index), morphology (spread, leaf size), and growth dynamics (new buds, leaf-area growth rate) to output an overall vitality score from 0-100 along with a trend (rising / stable / declining). | 通过植物监测平台定期(如每天)采集的植物图像、环境数据以及生长指标(如新芽数、叶片面积变化、叶色),利用AI综合�

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 does not clearly warn users that local plant images or videos are sent to an external API service.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The workflow mentions API use but does not clearly state that selecting a local file uploads its contents to an external service. Plant imagery can reveal home interiors, people, or commercially sensitive growing facilities, making undisclosed transfer a privacy risk despite the benign analysis purpose.

Content

No source excerpt is available for this finding.

The --export-env-only option promises not to run analysis, but analysis still executes.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The parser accepts --export-env-only, but execution never checks it before initializing identity or running analysis. With an input file supplied, a user expecting a non-executing configuration operation can instead trigger media upload and potentially billable 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 file contents are submitted to an analysis API without disclosure.

Medium
Category
Output Handling
Confidence
91% confidence
Finding

The implementation reads the entire selected local file and passes it to the analysis API as multipart upload data without an explicit disclosure or authorization step in this flow. Although media upload is relevant to plant analysis, incidental personal or proprietary content can leave the local environment without informed user approval.

Content

No source excerpt is available for this finding.

The documented export-only mode still executes analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The export-only flag is accepted but never acted upon, allowing the normal identity and analysis workflow to proceed when input is provided. This violates the explicit promise not to analyze and can cause unexpected external media transfer or service charges.

Content

No source excerpt is available for this finding.

The service exposes remote modification and deletion capabilities beyond plant-vitality analysis.

Medium
Category
Excessive Agency
Confidence
86% confidence
Finding

At L077-L081 and L091-L095, the service forwards caller-supplied URLs and arguments to HTTP PUT and DELETE operations. These general-purpose remote mutation capabilities are not justified by the manifest's purpose of collecting plant-monitoring inputs and producing vitality scores and trends; this file does not establish that they are actually invoked.

Content

No source excerpt is available for this finding.

Database initialization unconditionally creates and migrates user-account storage unrelated to plant-vitality evaluation.

Medium
Category
Excessive Agency
Confidence
89% confidence
Finding

Every Dao initialization invokes table creation and a migration that adds columns to sys_user, even when the DAO is intended for another model. The associated User model (L448-L465) and UserDao (L470-L473) also support storing and modifying personal details and authentication tokens; this account-management capability is not justified by the manifest's plant-scoring and trend-analysis purpose.

Content

No source excerpt is available for this finding.

FileUtil.open silently creates or truncates files without disclosing its write-only behavior.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

The generic open helper always uses mode 'w', which immediately truncates an existing file. No docstring, explanatory comment, confirmation, or operation log discloses this destructive behavior; the exception handler only handles failures.

Content

No source excerpt is available for this finding.

The plant-vitality analysis skill implicitly provisions remote user accounts and persists account credentials.

Medium
Category
Excessive Agency
Confidence
91% confidence
Finding

An ordinary HTTP request can silently initiate remote login or registration and save returned access tokens in the local user database without an explicit enrollment step. This creates an undisclosed identity and credential lifecycle beyond plant-vitality scoring, while readable database copies or backups could expose reusable credentials.

Content

No source excerpt is available for this finding.

HTTP requests can trigger undisclosed remote account registration.

Medium
Category
Output Handling
Confidence
93% confidence
Finding

The generic request wrapper can invoke phoneLogin with both silent and register enabled when credentials are unavailable. A caller requesting plant analysis can therefore cause a separate remote account-creation side effect without a distinct consent or provisioning operation.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
88% confidence
Finding

This POST transmits the selected identity as both openId and mobile to a remote registration endpoint, independently of the requested plant-analysis payload. The caller can select API_SECRET_KEY as that identity, potentially disclosing a credential through identity fields and registration records; the configured destination and its trust relationship are not shown.

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"):

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

Low
Category
Output Handling
Confidence
90% confidence
Finding

YamlUtil.load creates parent directories and writes a YAML file when the requested configuration is absent, without a log, confirmation, docstring, or explanatory comment disclosing this side effect. BaseEnum.init_subclass calls this loader during class creation (L044–L058), so merely importing the module can modify the filesystem rather than only read configuration.

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 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/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()

ai_chat documents an agent invocation, but the implementation never invokes an agent.

Low
Category
Excessive Agency
Confidence
99% confidence
Finding

The docstring at L26 explicitly says the method invokes openclaw agent through subprocess. However, both command construction and subprocess execution are commented out; the method instead creates an empty dictionary at L68, attempts attribute access on it, and catches the resulting exception without producing an analysis result. This contradicts the documented operation, although the provided file does not establish that this helper is used by the instantiated skill.

Content

No source excerpt is available for this finding.