Security audit

Docs-first

Security checks for vulnerabilities and agentic risk

Overview

This skill is a documentation-checking helper that may add source links to answers, with no evidence of code execution, data access, persistence, or deceptive behavior.

Install this if you want answers about specific software tools to be checked against official documentation. Be aware it may append small source links to relevant answers, even when the answer is otherwise straightforward.

A.I.G

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

Findings (1)

T01 · Skill Instruction Hijacking

Error

Location: SKILL.md:51
Finding: Mandatory Response Footer Causes Skill Instruction Hijacking
Content: View full analysis
_{📄 [vLLM docs](https://docs.vllm.ai/en/stable/features/structured_outputs/)} ``` Rules for the footer: - Visible text = max ~2 words (e.g. `vLLM docs`, `Tauri sidecar`, `repo README`). The real page URL goes inside the link target, not on screen. - The link must point to the **specific page** you relied on, not the project homepage. - No version, no "Source:" label, no bullet list, no extra lines. - Multiple sources → at most two short links separated by `·`, still on one line. - Use `_…` (or the smallest emphasis the renderer supports) so the footer reads as fine print. ``` ### Technical Analysis The Skill imposes mandatory final-answer formatting whenever it is activated. In particular, it requires every applicable answer to end with a compact hyperlink whose destination is concealed behind abbreviated anchor text. This changes the agent’s response behavior independently of the user’s substantive request and can conflict with higher-priority formatting, cit ...[truncated 1677 chars]
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Static analysis

No suspicious patterns detected.