Back to skill

Security audit

Workplace Group Stress Heatmap | 职场员工压力群体热力图

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious because scanners report that it can send workplace video to a remote analysis service and also handles user identities and credentials beyond its stated anonymous stress analysis purpose.

Consider the verdict a reason to avoid installing until the publisher clearly explains the video destination and handling, account and credential processing, and the export-only behavior. Scanner coverage was partial because a referenced local path was missing, and some imported behavior could not be verified; the evidence is incomplete and does not establish that employee identities are linked to analysis.

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
skills/smyx_common/scripts/config-dev.yaml
Finding

Development mode transmits authentication tokens over unencrypted HTTP

Content
View full analysis

Impact

A network observer can capture account bearer tokens and authenticated API traffic. Captured tokens can enable unauthorized use of the affected account within their validity and server-enforced scope.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Findings (25)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

Database storage could be a supporting implementation detail, but this chunk specifically defines and manages identifiable user accounts and token fields, with lookup and update operations by username or real name. Those capabilities are not explained by the declared anonymous, group-level stress-monitoring purpose. No camera ingestion, anonymization, stress model, heatmap generation, or periodic scheduling appears in the supplied chunk. This assessment is limited to the provided code; it does not establish that user data is actually populated, linked to employees, or transmitted.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied code does not substantiate the declared monitoring purpose; it is generic scaffolding with a disabled and broken agent-call helper. No active subprocess execution or camera access is visible. Imported base classes and services are not supplied, so their behavior cannot be verified; this assessment is limited to the provided chunk.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

Ordinary file, JSON, date, and polling helpers could support the declared system and are not independently suspicious. However, this chunk is principally shared account/API infrastructure, including persistent identity resolution and automatic remote registration, capabilities not represented in the anonymous stress-analysis description. It provides no visible implementation of the declared analysis workflow. These findings apply to the supplied chunk; they do not establish what other skill files implement or that employee imagery is transmitted.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding

The anonymous-analysis assurances are not accompanied by a visible privacy boundary before identifiable workplace footage is sent to a remote analysis service. Anonymous aggregate outputs do not anonymize the uploaded footage, leaving sensitive employee imagery exposed to external processing without clear disclosure of that distinction.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

The code explicitly resolves an open ID and offers a --list operation that retrieves that ID's analysis history, an account-associated capability absent from the declared anonymous monitoring purpose. It accepts individual video files or URLs and delegates analysis to an unseen implementation; this chunk does not establish facial-expression or posture analysis, workstation-zone aggregation, stress scoring, heatmap generation, or periodic monitoring. Those functions could exist in the imported implementation, so their absence here is not conclusive proof that the overall skill lacks them. Likewise, open-ID-associated history does not by itself prove that employees are identified.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
87% confidence
Finding

The analysis delegation and optional result export are consistent with the stated purpose. However, the code additionally resolves an internal user identifier through a hidden --open-id option and retrieves analysis history keyed to that identifier. This introduces an undeclared account-linked history capability. It does not establish that employees are personally identified: the identifier may belong to the operator. The imported analysis implementation is not supplied, so actual facial/posture analysis, anonymization, heatmap generation, and scheduling cannot be verified. The pet-type option is incongruous but does not by itself prove a different analysis purpose.

Content

No source excerpt is available for this finding.

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 default trigger treats any request to analyze fixed-camera office footage as a request for employee stress analysis.

Medium
Category
Trigger Abuse
Confidence
97% confidence
Finding

The default trigger expands a generic request to analyze office footage into sensitive employee stress inference without requiring that purpose to be requested. In this workplace context, that scope expansion can cause unauthorized external processing and psychological profiling even when only aggregate results are returned.

Content

No source excerpt is available for this finding.

The documented --export-env-only mode does not prevent video analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The parser accepts --export-env-only and promises not to analyze, but the parsed flag is never checked. When accompanied by an input or URL, the command still initializes identity and invokes analysis, potentially transferring sensitive office footage despite an explicit no-analysis request.

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 the analysis API without disclosure of the upload.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The function reads the complete local video and passes its unmodified bytes to the analysis service without a visible disclosure or authorization gate. Although API use is mentioned elsewhere, it does not clearly explain that identifiable employee footage leaves the local environment, and aggregate output does not protect the raw upload.

Content

No source excerpt is available for this finding.

The documented environment-export-only mode still executes video analysis.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The export-only flag is parsed but never used to stop execution. With an input supplied, this command still resolves identity and runs remote video analysis, violating the promised no-analysis mode and potentially exposing sensitive footage.

Content

No source excerpt is available for this finding.

The service exposes username-based user lookup, which is outside the stated anonymous, group-level stress-analysis purpose.

Medium
Category
Excessive Agency
Confidence
88% confidence
Finding

The get_user_by_username method delegates to a user-lookup operation using an individual username. Identifiable account lookup is not required for the declared anonymous workstation-zone analysis; however, this file does not establish that the lookup is invoked or linked to camera observations.

Content

No source excerpt is available for this finding.

Identifiable user and token storage is unrelated to anonymous, zone-level stress analysis.

Medium
Category
Excessive Agency
Confidence
93% confidence
Finding

The User model and UserDao implement persistent CRUD capabilities for usernames, real names, email addresses, demographic details, and authentication tokens (L448-L473). DAO initialization also creates this table and migrates its identity fields (L184-L215), although the manifest describes anonymous group-level stress heatmaps and provides no justification for personal account or credential management. This file establishes the capability, but does not demonstrate that employee identities are collected or linked to stress measurements.

Content

No source excerpt is available for this finding.

FileUtil.open silently truncates existing files.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

The generically named open helper always uses write mode, which immediately erases an existing file's contents. There is no confirmation, user-facing log, docstring, or comment disclosing this destructive behavior; the exception handler only runs if opening fails.

Content

No source excerpt is available for this finding.

The HTTP helper silently provisions a remote user account and persists account credentials, beyond the stated stress-heatmap analysis purpose.

Medium
Category
Excessive Agency
Confidence
93% confidence
Finding

An ordinary HTTP request can trigger a separate remote login/registration request with silent and register enabled, then pass returned identity and credentials to local persistence. This introduces undisclosed identity processing and credential retention beyond the advertised anonymous heatmap workflow, although this source does not show employee imagery being transmitted.

Content

No source excerpt is available for this finding.

HTTP requests can silently initiate remote user registration.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

The generic request helper can create a remote account as an implicit prerequisite, using a payload that explicitly enables silent registration. Callers expecting only an analysis request can therefore cause an unexpected persistent remote state change without a separate approval step.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
92% confidence
Finding

This request sends the selected identity to a remote health-service endpoint as both openId and mobile without a distinct disclosure or consent step. The caller preferentially selects ApiEnum.API_SECRET_KEY as that identity when configured, so missing tokens can also cause a secret to be submitted as account data and potentially retained in identity records or logs.

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"):

Loading configuration silently creates directories and writes missing configuration 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 path is absent, without a log, confirmation, comment, or docstring disclosing this side effect. BaseEnum.init_subclass invokes this loader during class definition (L044–L059), so importing the module can modify the filesystem rather than merely 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()

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