Back to skill

Security audit

translate-polish

Security checks for vulnerabilities and agentic risk

Overview

This translation skill is mostly straightforward, but it needs review because its learning script stores cross-session notes and preferences and can be pointed at other local skill directories.

Review before installing. Use the translation and glossary-checking pieces only if you are comfortable with local file reads. Do not record client, legal, medical, personal, or confidential content in learner notes or preferences, and avoid running learner.py against any directory other than this skill's own directory. Remove learned_patterns.json when you no longer want the stored history retained.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (11)

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

This mismatch is more security-relevant because the skill claims to perform translation-related tasks while also documenting persistent recording of user notes, failures, and patterns across sessions. Users may provide sensitive legal, medical, business, or personal text under the assumption of a translation-only workflow, while the skill persists metadata and notes that could contain confidential information.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

This mismatch is more security-relevant because the skill claims to perform translation-related tasks while also documenting persistent recording of user notes, failures, and patterns across sessions. Users may provide sensitive legal, medical, business, or personal text under the assumption of a translation-only workflow, while the skill persists metadata and notes that could contain confidential information.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The module is explicitly designed as a generic cross-skill persistence and learning component for arbitrary WorkBuddy skills, which exceeds the declared translation/polishing scope. In a skill package, undisclosed generic stateful behavior increases attack surface because it can collect and retain operational data beyond what users would reasonably expect from a translation workflow.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

resolve_skill_dir accepts any existing directory path and also resolves arbitrary skill names under ~/.workbuddy/skills, allowing this module to read/write learned_patterns.json for other skills. That breaks isolation between skills and enables unauthorized tampering with or harvesting of usage history and preferences across the local skill environment.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The manifest description is written as a Chinese-only instruction for a translation workflow, and the skill content consistently assumes Chinese as the interface language. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless a justified locale restriction is clearly documented.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill introduces a cross-session self-learning system that records task outcomes and free-form notes, creating a persistent data retention channel in a context likely to involve sensitive documents. Translation tasks often include confidential contracts, medical content, or proprietary technical text, so retaining notes and error patterns can leak or accumulate sensitive information over time.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

Persistent recording of translation notes and error history is not clearly necessary for the stated purpose and increases privacy and confidentiality risk without strong justification. Because users may include client names, legal issues, medical conditions, or internal project details in notes, the retained history can become a sensitive corpus vulnerable to accidental disclosure or misuse.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module docstring and all user-facing CLI messages are written only in Chinese, and the interface expects Chinese-formatted term pairs like "源:目标". This imposes a specific language/locale on users without offering a choice or documenting a justified region-specific limitation, which matches the policy violation criteria for language/locale constraints.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The code persists preferences, operation history, notes, and error metadata to disk without any user-facing disclosure or consent mechanism. Because notes and preferences may contain sensitive business or personal content in a translation skill, silent retention creates privacy and compliance risk even if there is no immediate remote exfiltration.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
97% confidence
Finding

The natural-language documentation and user-facing strings are entirely in Chinese, which imposes a specific language on users without any opt-in or alternative. Under the language/locale policy, this is a policy concern unless the tool explicitly offers language choice or documents a justified locale restriction.

Content

No source excerpt is available for this finding.