Security audit
Security checks for vulnerabilities and agentic risk
This skill does perform cloud video analysis, but it also silently creates or reuses an identity, registers/logs in with a remote service, and stores tokens locally without clear user-facing control.
Install only if you are comfortable with pet videos or video URLs being sent to lifeemergence.com services, cloud report history being tied to an automatically managed identity, and local storage of account tokens in the workspace data directory. Avoid sensitive footage unless the publisher clarifies consent, retention, account creation, and deletion controls.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
The reported code accesses workspace files, generates persistent default user identities, stores authentication tokens, performs authenticated HTTP requests, and may auto-register or log in users via a remote API. These are significant hidden behaviors for a skill advertised as simple pet video analysis, creating risks of secret account creation, credential misuse, persistent tracking, and expanded backend access.
YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).
---
name: "smyx-pet-drying-box-heat-stress-analysis"
description: "Triggers when a user provides a pet drying box area video URL or file for analysis; supports local video uploads or network URLs to call server-side APIs for pet heat stress signal detection, analyzing open-mouth panting intensity, tongue color (pink/cyanotic), and body movement frequency to identify early heat stress signals, outputting risk levels and supporting auto-cooling or stopping drying. Application scenarios: pet drying boxes, pet grooming stores, pet hospitals. Development reason: prevent heatstroke and improve safety. | 当用户提供宠物烘干箱区域的视频URL
The HTTP helper can silently call an external /sys/phoneLogin endpoint with register=1 and silent=1 to auto-register/login users, then cache returned token and openToken. For a pet drying-box analysis skill, this is unrelated privileged behavior that can transmit identifiers off-platform and create remote accounts without meaningful user action, increasing privacy, account-abuse, and hidden data-flow risks.
The skill advertises behaviors that require shell, network, file, and environment access, but the manifest does not explicitly declare or constrain those capabilities. This creates an overbroad and opaque trust boundary, making it harder to audit what the skill may access and increasing the chance of unintended privileged execution.
The skill description does not clearly warn users that uploaded local videos or supplied URLs are transmitted to server-side/cloud APIs for analysis and history lookup. This lack of transparency undermines informed consent and may expose sensitive media, metadata, or internal URLs unexpectedly.
The trigger and usage guidance are primarily specified in Chinese keyword phrases and examples, with no indication that users may choose another language for invocation or results. This can constitute a language/locale policy issue because the skill appears to privilege a specific language without explicit opt-in or justification as a region-specific tool.
The manifest says users can provide local files or network URLs and that the skill calls server-side APIs for analysis. However, analyze_video accepts api_url, api_key, and output_level but does not use them, instead collapsing input to input_path or url and calling skill.get_output_analysis(input_path) unconditionally. This creates a semantic mismatch between the advertised configurable API-based behavior and the actual implementation exposed here.
The tool performs identity resolution through a hidden --open-id parameter and suppresses it from help output, reducing transparency around how user identity is selected. In a skill that can retrieve prior analysis lists, hidden identity selection increases the risk of unauthorized access or user confusion if callers can supply or influence open_id, especially because there is no visible validation or disclosure in this entrypoint.
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.
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
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.
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
The skill reads arbitrary local file contents into memory and sends them to a server-side analysis API, but this code contains no user-facing disclosure, confirmation, or consent checkpoint before exfiltration. In an agent setting, that creates a real privacy and data-handling risk because a user may provide a local path without understanding the file will be uploaded off-device.
The script accepts a local file path or remote video URL and forwards it to backend analysis via skill.get_output_analysis(...) without any user-facing notice, confirmation, or disclosure that data may be transmitted off-device. In this skill’s context, videos from pet drying boxes may contain sensitive operational footage, customer environments, or network-retrieved content, so silent transmission creates a real privacy and data-handling risk even if the functionality is intended.
This service exposes broad generic HTTP and CRUD wrappers (GET/POST/PUT/DELETE, add/edit/delete/list/page) that are not constrained to the declared pet heat-stress video analysis purpose. In a skill that should only submit video inputs to specific analysis endpoints, these primitives enlarge the attack surface and can be repurposed to access or modify unrelated backend resources if other parts of the skill pass attacker-controlled URLs or parameters.
The username lookup capability is unrelated to pet drying box video analysis and introduces unnecessary access to identity-related functionality. Even if harmless in isolation, surplus account-lookup features can enable user enumeration, privacy exposure, or chaining with other overbroad API methods in a context that should only process media for heat-stress detection.
This code opens a SQLite database, creates tables, and performs schema-altering migrations automatically during DAO initialization. Although comments describe the behavior for developers, there is no user-facing disclosure, confirmation, or warning before modifying persistent local data.
The file implements generic local user/account persistence unrelated to the declared pet drying-box heat-stress video analysis purpose. This kind of capability expansion increases the data-collection and attack surface of the skill, especially when coupled with token storage and record mutation methods that are not necessary for the advertised function.
The delete method permanently removes a database record and commits the change immediately. The method has an internal docstring, but no visible confirmation prompt, user disclosure, or external documentation warning about this destructive operation.
Detected: suspicious.install_untrusted_source