ClawHub
Back to skill

Security audit

Presence Recall

Security checks for vulnerabilities and agentic risk

Overview

This is a simple reflection/tone skill with no executable code, persistence, network access, credentials, or system-level authority.

Install this only if you want a tone-calibration reminder that may slow responses and make them more reflective. Avoid enabling it for workflows that require fast, factual, policy-bound, or safety-critical answers unless the agent is instructed to prioritize those requirements over this style guidance.

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

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (2)

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The skill explicitly encourages the agent to answer from internal feelings and to infer user intent beyond the user's words (for example, assuming what the user 'just wants'). This can produce unsupported personal inferences, anthropomorphic fabrication, and responses grounded in speculation rather than evidence, which is risky in safety-sensitive or factual interactions.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The invocation guidance is intentionally broad and subjective ('anytime', 'when you notice', 'feels empty'), which makes the skill easy to trigger in unrelated contexts. Because it changes the agent's response style based on internal state rather than task requirements, it can unpredictably interfere with normal instruction-following and increase the chance of off-task or policy-inconsistent behavior.

Static analysis

No suspicious patterns detected.