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
Security audit
Context Compressor
Security checks for vulnerabilities and agentic risk
Overview
This is a local Python text-compression skill that reads user-selected text files or stdin and prints compressed summaries or extracted facts, with no evidence of network access, persistence, hidden execution, or destructive behavior.
Install only if you are comfortable giving the skill local files you explicitly choose to compress or analyze. Avoid running it on secrets or highly sensitive logs unless you intend the resulting summaries, diffs, URLs, paths, or snippets to appear in terminal output, and review compressed text before using it to replace original conversation context.
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
- MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (6)
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 77% confidence
- Finding
The skill title and introductory content begin in Chinese and present bilingual labeling, but there is no indication that language choice is user-selectable or that Chinese is required for a region-specific purpose. This may conflict with a language/locale policy if users are not given an explicit choice.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 82% confidence
- Finding
This markdown file describes replacing older conversation content with compressed output, which can affect user data fidelity, but the usage section does not explicitly warn users before suggesting substitution. Although lossy behavior is mentioned later in the limitations section, the operational example that encourages replacement would benefit from a clear warning near the action itself.
- Content
Description-Behavior Mismatch
Low
- Category
- Not specified by scanner
- Confidence
- 87% confidence
- Finding
The manifest describes a skill focused on intelligently compressing context, but the module documentation and code expose separate capabilities for diffing files, extracting structured information, and computing text statistics. These are adjacent text-analysis functions, but they go beyond straightforward compression and are not reflected in the stated skill description.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 82% confidence
- Finding
This code reads the full contents of a user-specified file and then prints compressed output to stdout. For a context-compression tool, this behavior may expose sensitive file contents in logs, terminals, or calling pipelines, but there is no explicit warning or disclosure in the code about handling potentially sensitive data.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 85% confidence
- Finding
The diff command reads two files and prints changed lines, and the extract command reads input and emits structured data such as URLs and file paths. These are user-data-affecting operations that can surface sensitive content, yet there is no visible warning, prompt, or disclosure to the user in the command implementation.
- Content
Static analysis
No suspicious patterns detected.