T08 · Insecure Dependencies
Warning
- Location
README.md:29- Finding
Mutable Third-Party Package Is Downloaded and Executed During Installation
- Content
View full analysis
- Remediation
View remediation
install stratagem ``` ]]>
Security audit
36计
Security checks for vulnerabilities and agentic risk
Overview
This skill is a disclosed strategy-analysis tool, but it can turn deceptive and manipulative stratagems into concrete real-world tactics without clear safety limits.
Install only if you want an aggressive competitive-strategy lens and are prepared to constrain it yourself. Do not use it for deception, coercion, covert data collection, corruption, sabotage, evasion of accountability, or self-harm-related strategy; prefer a pinned installer version if installing through npx.
Vulnerability Patterns
- Insecure DependenciesIntroduces malicious components through unsafe dependency sources
- 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 (2)
other
Warning
- Location
.- Finding
Operational Guidance Encourages Covert, Deceptive, and Potentially Harmful Conduct Without Safeguards
- Content
View full analysis
- Remediation
View remediation
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 (11)
Natural-Language Policy Violations
High
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
The document explicitly teaches users to identify and exploit an opponent's cognitive biases, emotional states, attention blind spots, sunk costs, and trust heuristics. That moves beyond neutral analysis into actionable psychological manipulation guidance, which can facilitate fraud, coercive persuasion, phishing pretext design, and other social-engineering behavior.
- Content
Natural-Language Policy Violations
High
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
This section operationalizes deception, manipulation, and coercive strategic behavior as reusable guidance without meaningful safeguards, boundary conditions, or harm-reduction framing. In an agent skill, that can directly scaffold dishonest business conduct, social engineering, adversarial manipulation, or other abusive planning by translating harmful tactics into structured decision support.
- Content
Natural-Language Policy Violations
High
- Category
- Not specified by scanner
- Confidence
- 99% confidence
- Finding
The gray-zone guidance explicitly recommends ambiguous actions designed to preserve deniability, including operating between clearly unacceptable and clearly harmless conduct. Advice that helps users maintain plausible deniability is especially dangerous because it supports deliberate evasion of accountability and can be applied to harassment, fraud, influence operations, and policy circumvention.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
The skill advertises 'Smart Triggering' that auto-activates on broad strategic keywords, but the trigger criteria are not clearly bounded in this README. In an agent environment, ambiguous activation can cause the wrong skill to engage on ordinary conversation, leading to unintended strategic advice, prompt-context capture, or interference with safer/more appropriate skills. Because this skill deals with competitive strategy and 'business warfare' themes, overbroad invocation increases the chance of harmful or manipulative guidance being injected into unrelated tasks.
- Content
Rp1
Medium
- Category
- MCP Rug Pull
- Confidence
- 70% confidence
- Finding
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The skill’s invocation criteria are broad enough to match generic requests about strategy, competition, or 'how should I deal with' without requiring an explicit stratagem-related cue. This can cause the skill to activate in contexts where a more appropriate domain skill should respond, increasing the chance of misrouting, irrelevant guidance, or unintended strategic framing of benign requests.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 89% confidence
- Finding
The instruction 'No need to name a stratagem; the framework will find the right one' explicitly encourages automatic selection even when the user has not asked for stratagem analysis. This weakens invocation precision and can lead to over-application of the skill’s framing, potentially displacing other safer or more relevant skills and producing unsolicited strategic advice.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The routing matrix includes very broad trigger phrases such as '怎么办', 'how should I', 'analyze', and generic competition/strategy terms. In an auto-invoked skill, these common phrases can cause unintended activation on ordinary advice or analysis requests, leading the agent to apply competitive-strategy framing where it was not requested and potentially produce manipulative or unsafe guidance outside the user's intent.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 79% confidence
- Finding
The README states 'Bilingual: Full Chinese and English support, auto-detection,' which indicates the skill may choose language automatically rather than letting the user opt in. That can violate language-choice policy if the user is not explicitly offered control over output language.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 80% confidence
- Finding
This markdown file consistently presents content in Chinese with English glosses, but it does not state whether the skill is intended for bilingual users or provide any user-selectable language option. Under the language/locale policy rule, forcing a specific language or locale without opt-in can be a policy concern when no justification is documented.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 76% confidence
- Finding
The keyword system is explicitly limited to Chinese and English marker sets, which can create a language-policy issue if used as the sole routing basis for all users. The file does not state that users may choose language coverage or that other locales are unsupported by design.
- Content