Security audit

iq-core

Security checks for vulnerabilities and agentic risk

Overview

This is a reasoning-guidance skill with a small local logging helper; its behavior is disclosed and aligned with its stated purpose.

Install this if you want the agent to apply more structured reasoning on complex tasks. Be aware it may be invoked broadly for difficult questions, and avoid putting secrets or sensitive personal data into the optional reasoning log helper.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
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 (7)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 77% confidence
Finding: The skill references executable capability via `scripts/reasoning_log.py` and includes behavior that can read environment context and write logs or evaluation artifacts, yet no explicit permissions are declared. This creates a transparency and policy-bypass risk: an agent may use file-write or environment-derived data without the caller having clear permission boundaries or review expectations.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The README instructs users to 'use when the conditions in the description are met' without defining concrete trigger criteria or boundaries. In a reasoning-activation skill, this ambiguity can cause over-broad invocation, inconsistent behavior, or accidental application in contexts where stronger safeguards or narrower instructions are needed.

Vague Triggers

Medium

Confidence: 87% confidence
Finding: The description positions the skill for broadly defined 'high-level reasoning' tasks, which can match a large fraction of normal user requests. Overbroad activation increases the chance that this skill's instructions override or steer unrelated tasks, causing unreviewed behavioral changes and accidental invocation of auxiliary tooling.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger engine says to activate whenever the request 'matches the description above,' which is subjective and lacks objective boundaries. Ambiguous activation logic is dangerous because it lets the skill self-justify invocation in many contexts, reducing predictability and making prompt-scope control harder.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: Instructing the agent to invoke this skill for 'any task requiring high-level reasoning' effectively expands the skill into a near-global policy layer. In agent systems, broad self-activation can crowd out safer, narrower skills and alter decision-making across many tasks without clear authorization.

Vague Triggers

Medium

Confidence: 78% confidence
Finding: The instruction to invoke the skill for 'any task requiring high-level reasoning' is overly broad and can cause unnecessary activation across many ordinary tasks. In an agent system, broad activation criteria can expand the influence of the skill beyond intended scope, increasing prompt-surface area and making behavior less predictable, especially when the skill modifies reasoning style or tool usage expectations.

Vague Triggers

Low

Confidence: 72% confidence
Finding: The description's trigger phrase 'activating high-level reasoning' is ambiguous and underspecified, which can lead to inconsistent or excessive invocation. While not directly malicious, vague activation semantics in shared agent environments can cause routing errors, unintended policy overlap, and reduced operator control over when this skill governs behavior.

Static analysis

No suspicious patterns detected.