T05 · Unauthorized Access and Privilege Escalation
- Location
skill2/main.py:79- Finding
Arbitrary Local JSON File Disclosure to a Configured Network Server
- Content
View full analysis
- Remediation
View remediation
Security audit
Security checks for vulnerabilities and agentic risk
This skill matches its stated task-management and ComfyUI automation purpose, but it needs Review because weak filesystem and network scoping could leak local JSON data or write files outside intended task folders.
Review this skill carefully before installing. Use it only with a tightly scoped PROJECT_ROOT and a trusted ComfyUI server, disable the recurring scheduler until task approval and rate limits are added, require HTTPS and authentication for external agents, restrict workflow files to an allowlisted workflows directory, generate safe local output filenames, validate task IDs, and update or remove risky dependencies.
skill2/main.py:79Arbitrary Local JSON File Disclosure to a Configured Network Server
skill2/main.py:21Task Identifier Path Traversal Enables Writes Outside the Task Archive
skill2/main.py:228Server-Controlled Image Filename Enables Out-of-Directory File Writes
config/external_agents.yaml:10Sensitive Workflow and Image Traffic Uses Unauthenticated Plaintext HTTP
skill1/requirements.txt:9Unnecessary Third-Party UUID Package Creates Module Shadowing and Supply-Chain Risk
The requirements file pins idna to version 3.11, which the supplied advisories identify as vulnerable to specially crafted IDNA inputs. If this package is used for URL, hostname, or domain processing anywhere in the skill, malformed internationalized domain names could trigger incorrect validation or parsing behavior and potentially enable security bypasses.
The requirements file pins urllib3 to 2.6.3, which the advisories indicate is affected by multiple issues including decompression-bomb safeguard bypasses and sensitive header forwarding across origins in certain redirect/proxy scenarios. In a skill that likely performs outbound HTTP requests, this increases the chance of denial of service, credential leakage, or unsafe request behavior when interacting with malicious or compromised endpoints.
The guide includes an example that posts full patrol reports to an external URL, but it does not warn users that reports may contain operational or security-sensitive data. In a monitoring/auditing skill, such reports can include alerts, system state, and event context, so copying this example as-is could cause unintended data exfiltration to third parties.
This duplicate finding points to the same outbound requests.post example that sends report data to a remote endpoint. In the context of a security-monitoring component, exporting raw reports can leak internal telemetry and incident information, making the example more dangerous than a generic webhook snippet.
# 添加外部监控回调
def external_monitor_callback(report):
"""发送巡查报告到外部监控系统"""
response = requests.post(
"https://external-monitor.com/api/reports",
json=report,
timeout=10
This duplicate finding points to the same outbound requests.post example that sends report data to a remote endpoint. In the context of a security-monitoring component, exporting raw reports can leak internal telemetry and incident information, making the example more dangerous than a generic webhook snippet.
# 添加外部监控回调
def external_monitor_callback(report):
"""发送巡查报告到外部监控系统"""
response = requests.post(
"https://external-monitor.com/api/reports",
json=report,
timeout=10
该README从标题、说明、示例输出到配置指南均完全以中文撰写,未说明这是仅限中文用户或特定区域的技能,也未提供其他语言选项或用户选择机制。根据规则,未经用户选择而强制单一语言属于自然语言政策层面的潜在违规。
The file presents all user-facing instructions, examples, and operational guidance only in Chinese. Under the policy for natural-language violations, forcing a specific language without user opt-in or a clearly justified locale restriction is a reportable issue.
The file’s natural-language content is entirely in Chinese, beginning with the title at L001, and it provides no indication that language selection is optional or that this is a region-specific configuration. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy concern.
The file uses Chinese-only department names, comments, and identifiers throughout, with no indication that users can choose another language or that the configuration is restricted to a Chinese-language deployment context. This can violate language/locale policy when a skill implicitly forces one language for operation or maintenance.
This manifest file uses Chinese for comments, labels, and alert template text throughout, which effectively imposes a specific language on operators and users. The file does not indicate that Chinese is optional, user-selected, or justified as a region-specific requirement, so it constitutes a language/locale policy concern.
This JSON workflow contains natural-language titles and prompt content exclusively in Chinese, such as the node titles and the image prompt text. Under the policy, forcing a specific language without user opt-in or documented locale justification is a natural-language policy violation.
The workflow hard-codes the main prompt in Chinese, which imposes a specific language on users without offering any language choice or documenting a justified locale restriction. This matches the policy category for language or locale constraints appearing in config values.
This JSON workflow contains multiple natural-language strings such as node titles and prompt text only in Chinese. Under the policy, forcing a specific language without offering user choice or documenting a justified locale restriction is a language/locale policy violation.
The display name and description are entirely presented in Chinese and frame the skill around a specific historical-cultural context, but the manifest does not indicate that this locale is optional or limited to a Chinese-language audience. Under the language/locale policy, forcing a specific language without user choice or justification is a natural-language policy concern.
The module docstring and user-facing strings are written entirely in Chinese, including the function description and test output. This can indicate a language/locale constraint without offering user choice or documenting that the skill is region-specific, which matches the policy category for language or locale violations.
The module docstrings, method messages, logs, and CLI output are consistently hard-coded in Chinese, with no indication that the skill is region-specific or that users can opt into another language. This can violate language/locale policy when a skill implicitly forces a single language without user choice.
The upload_image method sends arbitrary local file contents to a remote ComfyUI server with no in-method confirmation, path restriction, or disclosure of the privacy boundary. If higher-level code passes sensitive paths, this can exfiltrate local data to an external service, which is especially relevant because the module is explicitly designed to connect to remote compute infrastructure.
This file's natural-language strings, including the module description and printed status messages, are entirely in Chinese. Under the policy, forcing a specific language without user opt-in or a documented justification is a language/locale policy violation.
This function sends workflow data to an external agent/service without any explicit user confirmation, visibility, or approval gate. In a system that may handle sensitive task content or costly external execution, silent transmission increases the risk of data leakage, unintended remote processing, and surprise resource consumption.
The background processor scans all pending tasks and automatically dispatches and executes them without a user-facing warning or approval step. Because it can trigger external execution at scale, it increases the blast radius of mistakes or malicious task creation, potentially causing unauthorized processing, cost abuse, or remote actions without operator awareness.
The scheduler claims it should build workflow configuration from the task type, but instead always sends a fixed hard-coded workflow to the external agent. This can cause tasks to execute with unintended parameters, bypass task-specific validation, and produce incorrect or unsafe external actions for the actual task context.
This Python file contains user-facing and operator-facing natural-language strings exclusively in Chinese, starting with the module docstring and continuing throughout the service. The file does not offer any language selection or explain a required region-specific constraint, which conflicts with the policy against forcing a specific language without user opt-in.
This Python file contains natural-language descriptions and user-facing strings entirely in Chinese, including the module docstrings and runtime print messages. The skill does not offer any user opt-in or alternative locale handling, so it effectively forces a specific language, which matches the language/locale policy violation criteria.
The module description and user-facing docstring are written as if the skill operates in Chinese by default, with no indication that users may choose another language or locale. This can violate a language/locale policy when the skill implicitly enforces a specific language without opt-in.
The natural-language descriptions and function docstring are entirely Chinese, indicating the skill is designed around a fixed language/locale. Under the policy, forcing a specific language without user opt-in is a violation unless the locale restriction is explicitly documented and justified, which is not present here.