Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md porcentaje, un ORO o un veredicto: todo sale de `scripts/motor.mjs`. Si escribes una conclusión con
Security audit
analiza-tu-contenido-f100k
Security checks for vulnerabilities and agentic risk
Overview
The skill is mostly a coherent Instagram content-analysis workflow, but it includes under-scoped fallback collection paths that could be used on accounts beyond the stated owner-controlled use case.
Install only if you intend to use it for an account you control or have explicit consent to analyze. Prefer the Composio connected-account path, avoid the unofficial Instagram scraping recipe, and be aware the generated datos.json and HTML report may contain private business metrics and should be stored/shared carefully.
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
Vulnerability Patterns
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- 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 (11)
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md un número, cópialo del `resultado.json`. El validador de `render.mjs` rechaza lo que contradiga al motor.
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md - **formato** = el MOLDE, con el `nombre` exacto de `references/formatos.json` (186 de la Biblioteca). Si ninguno encaja, descríbelo con palabras de mecánica, n
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md - **formato** = el MOLDE, con el `nombre` exacto de `references/formatos.json` (186 de la Biblioteca). Si ninguno encaja, descríbelo con palabras de mecánica, n
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md - **formato** = el MOLDE, con el `nombre` exacto de `references/formatos.json` (186 de la Biblioteca). Si ninguno encaja, descríbelo con palabras de mecánica, n
Description-Behavior Mismatch
Medium
- Category
- Not specified by scanner
- Confidence
- 90% confidence
- Finding
The skill says it should only be used for the student’s own account with her statistics, but it also documents a fallback that can pull public Instagram data for any handle via Apify. That creates a privacy and authorization gap: an operator could analyze a third party’s account without consent, contrary to the stated restriction, and normalize collection of profile data outside the intended trust boundary.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 97% confidence
- Finding
The instruction 'Todo lo que lee va en tú (nunca vos)' imposes a specific language/locale variant on all output. This is a natural-language policy concern because it removes user choice about dialect/register and is not documented as a justified regional compliance requirement.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The code forces all numeric and date formatting to use the "es-CO" locale via toLocaleString and toLocaleDateString. This is a natural-language/locale policy issue because the file does not offer user opt-in, fallback, or any documented reason that the report must be rendered specifically for Colombian Spanish users.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
SQP-3 applies to all file types and covers natural-language language/locale policy violations. This JSON defines skill content fields and long-form instructional text exclusively in Spanish, with no indication that the skill is region-specific or that users can opt into that language.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
This code file contains extensive user-facing instructional text and labels exclusively in Spanish, such as the header comments and explanatory strings. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified.
- Content
Intent-Code Divergence
Low
- Category
- Not specified by scanner
- Confidence
- 82% confidence
- Finding
The documentation presents Apify as a limited fallback, but later normalizes direct GraphQL/API scraping of Instagram endpoints as a working acquisition method. Even if framed operationally, these instructions encourage collection through unofficial interfaces that may bypass platform controls, increase legal/compliance risk, and make it easier to fetch data outside properly authorized integrations.
- Content