ClawHub
Back to skill

Security audit

carrusel-del-radar-f100k

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Instagram carousel workflow with approval gates before spending and publishing, though users should understand it uses local profile data and connected posting tools.

Install only if you want this agent to use your local brand/profile materials, Gemini image generation, Telegram review messages, and a connected Composio Instagram account. Confirm the companion skills and Composio account are trusted, and keep the final publish phrase requirement in place.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (4)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation phrases are broad enough that ordinary user replies such as 'carrusel del radar', 'hazme un carrusel de [tema]', or responses to pending approvals could invoke the workflow unintentionally. Because this skill can trigger external actions, persist state, send Telegram messages, generate paid images, and eventually publish to Instagram, accidental invocation can cause unwanted processing and side effects even if final publication is gated behind approval.

Skill Enumeration

Medium
Category
Agent Snooping
Content
Lee `skills/carrusel-formatos-f100k/references/catalogo.md` y elige la ficha **por objetivo**
(alcance, guardados, seguidores o venta), igual que la skill. Mira la columna «¿Quién lo fabrica?»
de `skills/carrusel-formatos-f100k/SKILL.md`.

**En este flujo solo entran fichas que se fabrican sin que ella te mande nada:**
Confidence
80% confidence
Finding
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Natural-Language Policy Violations

Low
Confidence
98% confidence
Finding
The instruction "Tuteo, español neutro" imposes a fixed language/register policy. There is no indication that the user can choose another language or locale variant, nor that this restriction is justified as region-specific compliance.

Natural-Language Policy Violations

Low
Confidence
87% confidence
Finding
The description "ISO 8601 con zona de USER.md" implies the skill must use the timezone defined in USER.md rather than allowing user opt-in or selection. Under the policy rule, forcing a locale-related setting without explicit choice or documented justification is a natural-language policy concern.

Static analysis

No suspicious patterns detected.