Security audit

Work Productivity Tavily Web Workflow Helper

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-style workflow helper with broad routing language, but it does not contain code execution, credential handling, persistence, or destructive behavior.

Install only if you want a general Tavily or web-search workflow helper. Be aware that its broad keywords and implicit invocation may cause it to appear for ordinary web/search/API requests, so users or maintainers may want tighter activation rules.

A.I.G

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

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (18)

Vague Triggers

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The description contains broad terms like 'work-productivity', 'web', 'search', and 'implementation support' that can match many unrelated user requests, increasing the chance the skill activates outside its intended scope. Over-broad activation can cause the agent to apply the wrong workflow or instructions in contexts where they were not requested, which is a prompt-routing security and reliability issue.
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The trigger list includes generic keywords such as 'web', 'search', 'via', 'api', 'user', and 'alternative', which are common across many benign prompts. This makes accidental or adversarial invocation much easier, allowing the skill to intercept unrelated tasks and potentially override more appropriate handling.
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 93% confidence
Finding: 技能描述把启用条件定义得过于宽泛，只要用户提到若干常见主题或需要“流程、分析、实现支持”就可能触发，边界不清晰。这样的描述会让编排器或调用者难以准确判断适用场景，增加误用、过度调用和与其他技能竞争触发的概率。
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: 触发关键词包含非常宽泛的日常词汇，如“web”“user”“api”“alternative”等，容易与大量普通请求重叠，导致技能在并非需要该能力时被意外激活。误触发会让代理偏离用户真实意图，插入不相关流程或建议，并在多技能环境中放大错误路由与不必要的数据暴露风险。
Content: No source excerpt is available for this finding.

Vague Triggers

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The trigger sentence begins with a very broad phrase ('Help me') and then embeds generic requirement text, which increases the chance that unrelated user requests will accidentally match and invoke this skill. In an agent ecosystem, overbroad activation can route tasks to the wrong skill, causing unintended web-search-oriented behavior, confusion, or unsafe delegation in contexts the user did not intend.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The trigger sentence is extremely broad and can match ordinary user language rather than a narrowly scoped intent to use this specific skill. That increases the chance of accidental activation or prompt-routing to the skill in contexts where web/search actions, workflow generation, or related behaviors were not explicitly requested, which can lead to unintended tool use or over-collection of external information.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The invocation example describes a broad request pattern without clear scope, constraints, or confirmation requirements. In an agent ecosystem, vague invocation guidance can cause the system to apply the skill to loosely related requests, producing unintended searches, workflow recommendations, or configuration changes beyond what the user actually asked for.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger phrases are broad, generic, and include common workflow language such as 'help me' and 'I need a practical workflow', which can cause the skill to activate in unrelated conversations. In an agent ecosystem, overbroad invocation increases the chance of unintended routing, context capture, or execution of the wrong workflow, especially because this skill is associated with web/search-style tasks.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The example triggers are malformed, overly broad, and do not define clear invocation boundaries, so they normalize vague prompting rather than explicit skill selection. This increases the risk of unintended activation and makes it harder for downstream systems or users to distinguish when the skill should actually run.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 84% confidence
Finding: 该文件为 SKILL.zh-CN.md，整体内容仅以简体中文呈现，未见任何允许用户选择语言或说明仅面向特定中文区域用户的文字。根据规则，若技能强制特定语言而没有用户 opt-in 或明确合理的区域性约束，属于自然语言策略问题。
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: 示例触发句使用了“Help me”“I need a practical workflow”等高度泛化表达，和普通求助语句大量重叠，会进一步训练或暗示系统采用宽松匹配。虽然它本身不直接执行危险操作，但会显著增加技能被错误选中的机会，降低路由精度。
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The default prompt uses a very broad activation phrase tied to generic productivity and Tavily/web workflow language, which can overlap with ordinary user requests. In combination with skill routing, this can cause unintended invocation of the skill in contexts the user did not explicitly request, increasing the risk of prompt-surface expansion and accidental execution of skill-specific behavior.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: Enabling implicit invocation without tight activation constraints allows the platform to auto-select this skill based on loosely matching user language. Because the skill is framed around broad web/search/productivity workflows, this raises the chance of unauthorized or surprising invocation, which can expose users to unintended tool usage, incorrect task routing, or policy bypass opportunities through overly permissive activation.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The sentence 'I need a practical workflow for ...' is still too generic because many benign productivity requests could fit that pattern without actually requiring this skill. Ambiguous activation conditions make the router less precise, which is particularly risky for a skill centered on web/search workflows because it may be selected for unrelated tasks and produce irrelevant or externally connected actions.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 81% confidence
Finding: The user-facing text includes Chinese terms within otherwise English documentation, such as "Tavily 搜索-style workflows," without explaining language expectations or offering a language preference. This can create an implicit language constraint or inconsistent locale experience for users.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 72% confidence
Finding: This is a Chinese README, but substantial instruction content and trigger phrases are written in English, including the activation examples. That can impose a language expectation without explicitly giving users a language/locale option or explaining the intended audience constraint.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 82% confidence
Finding: The phrase 'Tavily 搜索-style workflows' introduces a Chinese term into the core description, suggesting a language-specific framing without any opt-in or explanation. For a general-purpose skill, this can create an implicit language/locale bias that is not documented or user-selectable.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 78% confidence
Finding: The natural-language description includes both English and Chinese ("搜索-style") without explaining whether language selection is user-driven or required for a specific region. This can conflict with language/locale policy expectations when no user opt-in or justification is provided.
Content: No source excerpt is available for this finding.