T09 · Insecure Skill Coding Practices
Warning
- Location
analyzer-prompt.md:24- Finding
Untrusted tender content is directly interpolated into privileged analysis prompts
- Content
View full analysis
- Remediation
View remediation
Security audit
Tender Analyzer Agent
Security checks for vulnerabilities and agentic risk
Overview
The skill is a coherent tender-analysis prompt package, but it needs review because it handles sensitive bid materials and includes under-scoped automatic learning and competitive-intelligence guidance.
Review this skill before installing. Use it only with documents you are allowed to share, redact confidential pricing, personal data, trade secrets, and non-public procurement details, and do not rely on its automatic learning or competitive-intelligence guidance unless sources are lawful, public, reviewed, and auditable.
Vulnerability Patterns
- Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
- 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
- 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)
T02 · Agent Memory Poisoning
Warning
- Location
learning-framework.md:57- Finding
Automatic external learning can persist untrusted information into the knowledge base
- Content
View full analysis
- Remediation
View remediation
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
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (14)
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The README explicitly encourages users to paste full tender and bid documents into the assistant, but provides no warning that those documents may contain confidential business information, pricing, personal data, or procurement-sensitive material. In a document-analysis skill, this omission materially increases the risk of unintended data disclosure because users are being prompted to share exactly the kind of sensitive content the tool is designed to process.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 91% confidence
- Finding
The file title and all operating instructions are written exclusively in Chinese, and the output template requires Chinese headings and labels. This imposes a specific language/locale on the interaction without any opt-in or alternative language option, which matches the policy category for language or locale constraints.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 91% confidence
- Finding
This markdown file presents all instructions and checklist items exclusively in Chinese, and nowhere indicates that the skill is intentionally limited to Chinese-speaking users or offers an opt-in language/locale choice. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The section on competitor intelligence gathering lists sensitive channels such as customer feedback, supplier information, and employee job-hopping information without any warning about privacy, legality, ethics, or limits on acceptable collection methods. In this context, the omission is dangerous because the document operationalizes competitive intelligence and may normalize improper surveillance, misuse of personal data, or acquisition of confidential business information.
- Content
Context-Inappropriate Capability
Medium
- Category
- Not specified by scanner
- Confidence
- 89% confidence
- Finding
The skill explicitly recommends collecting competitor intelligence from 'employee job-hopping information', which can incentivize gathering personal or non-public employment information for competitive advantage. In a bid/competition context, this creates privacy, confidentiality, and trade-secret risks because users may interpret it as approval to exploit insider knowledge or sensitive personnel movements.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 92% confidence
- Finding
The entire skill content, headings, examples, and output template are written only in Chinese, with no indication that users may choose another language or locale. Under the policy, a skill that effectively enforces a specific language without user opt-in is a natural-language policy violation unless the locale constraint is explicitly documented and justified.
- Content
Context-Inappropriate Capability
Medium
- Category
- Not specified by scanner
- Confidence
- 90% confidence
- Finding
The inclusion of '关系深浅' as a pricing factor steers users toward relationship-based influence rather than neutral, merit-based bid analysis. In a bidding/procurement context, this can encourage unfair or improper practices and normalize decisions based on personal connections instead of lawful evaluation criteria.
- Content
Context-Inappropriate Capability
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
Classifying customers as 'relationship-oriented' frames procurement strategy around influence and connections, which is especially risky in regulated bidding environments. This can facilitate unethical conduct, favoritism, or attempts to circumvent fair-competition rules by tailoring tactics to perceived susceptibility to relationship pressure.
- Content
Context-Inappropriate Capability
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The recommendation to '加强本地合作' as a relationship strategy moves from passive analysis into prescriptive influence-seeking behavior. In procurement settings, such advice can be used to cultivate improper local connections or favoritism, increasing corruption and bid-rigging risk.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 81% confidence
- Finding
The file’s natural-language instructions and structure are entirely in Chinese, and there is no indication that users can opt into another language or that the skill is limited to a Chinese-language or China-only deployment context. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 85% confidence
- Finding
The natural-language instructions and example prompts are exclusively in Chinese, which effectively constrains use to a single language without explicit user opt-in. Under the stated policy, language restrictions should either be optional or clearly documented as a justified locale-specific limitation.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 85% confidence
- Finding
The natural-language interface, headings, commands, and examples are presented entirely in Chinese, which effectively forces a specific language for use of the skill. The policy allows language constraints only when the skill offers user choice or clearly documents a justified locale-specific limitation, neither of which is present here.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 85% confidence
- Finding
SQP-3 applies to all file types and covers language or locale policy violations in natural-language content. This sample report forces a single language throughout the file and does not provide user opt-in, alternatives, or justification that the language restriction is required.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 92% confidence
- Finding
This markdown file presents all instructional content in a single language and does not indicate that the user can opt into Chinese or select another locale. Under the policy scope provided, forcing a specific language without user choice is a natural-language policy concern unless the locale limitation is explicitly justified.
- Content
Static analysis
No suspicious patterns detected.