Back to skill

Security audit

ticker-pipeline

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed financial research-report skill that uses market/web data and writes a report file, with no evidence of hidden execution, credential access, exfiltration, or destructive behavior.

Install only if you want a Chinese-language ticker research workflow that may call market/news/web tools, launch subagents, and create persistent Markdown reports in your workspace. Treat generated financial ratings as research support, not guaranteed investment advice.

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
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (9)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The trigger phrases are broad and do not document exclusion conditions, so the skill may activate in contexts where the user did not intend a full multi-agent pipeline. Because the skill can launch subagents, access external data sources, and write reports, accidental invocation increases the chance of unnecessary data access and unintended side effects.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill is presented primarily as an analysis pipeline, but it also performs persistent filesystem writes into the user's workspace. That side effect can surprise users, create unwanted artifacts, and overwrite or proliferate reports without an explicit consent step, especially if the skill is auto-invoked from broad trigger phrases.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill specifies automatic creation of report files in the user's workspace without an explicit warning or consent mechanism. Silent filesystem modification is dangerous because it can leak sensitive analysis into persistent storage, clutter the workspace, or overwrite existing files when users only expected an on-screen response.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The natural-language instructions are entirely in Chinese and define the agent role and task in Chinese without indicating that the user may choose another language. This creates a language/locale policy issue because it implicitly constrains interaction language without documented opt-in or justification.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The file presents the entire agent role and task specification exclusively in Chinese and does not indicate that users may choose another language. Under the policy rule, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The prompt explicitly permits fallback to search_web/tencent-news when the declared primary market source is unavailable, which contradicts the stated pipeline guarantee that missing core data should cause termination. This weakens data provenance controls and allows unstructured external content to influence a supposedly deterministic financial decision pipeline, increasing the risk of inconsistent, manipulated, stale, or hallucination-prone inputs.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The file is written entirely as a Chinese-only system prompt and requires specific Chinese output values such as regime and rating labels in the JSON schema. There is no indication that the user may choose another language or that the locale restriction is required for a documented region-specific use case.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The title, section headings, and all report text are written entirely in Chinese, which establishes a fixed output language for the skill. Under the policy, a skill should not force a specific language or locale unless it offers user choice or clearly documents a justified region-specific constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
79% confidence
Finding

The manifest description and the rest of the skill content are entirely Chinese and define fixed Chinese output artifacts and labels, but do not tell users they can choose another language. This can violate language or locale policy when a skill effectively forces one language without explicit opt-in or justification.

Content

No source excerpt is available for this finding.