T05 · Unauthorized Access and Privilege Escalation
Warning
- Location
SKILL.md:25- Finding
Excessive Tool Permissions Violate Least-Privilege Principles
- Content
View full analysis
- Remediation
View remediation
Security audit
can-free
Security checks for vulnerabilities and agentic risk
Overview
The skill appears to be a local hashing and logging helper, but it asks for broad command and file authority and includes unclear callback/API behavior.
Review before installing. Use this only in a tightly sandboxed environment, avoid sensitive inputs unless you know where logs are written, disable or ignore callback/API behavior unless explicitly needed, and do not run it with administrator privileges for normal use.
Vulnerability Patterns
- Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
- 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 (1)
Vulnerability Patterns
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)
Description-Behavior Mismatch
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The skill is presented as a local-only, self-evaluating content-addressing tool, but the documented interface includes a callback URL and later references external API behaviors. This mismatch can cause users or orchestrators to send data off-host under the false assumption that processing is purely local, creating privacy, policy, and trust-boundary risks.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 89% confidence
- Finding
The documentation exposes callback/API-style network behavior without a clear privacy or transmission warning. Users may provide sensitive content for hashing/logging while unaware that metadata or payloads could be sent to external services, which is especially risky because the skill markets itself as a local auditing utility.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The skill instructs users to append records to a local log file but does not clearly warn that it performs persistent local writes. In agent environments, silent file creation or modification can violate user expectations, overwrite sensitive locations if paths are configurable elsewhere, or create audit/privacy issues through retained data.
- Content
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The documentation says the free edition lacks endpoint validation, yet other sections discuss API authentication failures, rate limits, timeouts, and callback-driven remote interactions. These contradictory claims obscure the real attack surface and may lead operators to deploy the skill without appropriate egress controls, secrets handling, or monitoring.
- Content
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 91% confidence
- Finding
The skill repeatedly frames itself as local append-only logging with no external storage, but later advertises API integration and command execution as core capabilities. Combining undisclosed exec and external communication increases the risk of unintended code execution, data exfiltration, and operator misuse because the documented trust boundaries are inconsistent.
- Content
Static analysis
No suspicious patterns detected.