Back to skill

Security audit

Caricature Portrait Generator

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to generate caricature images as advertised, but it needs review because it handles an API token and sends it to a host that is not clearly disclosed in the main skill instructions.

Review this before installing. Use only a narrowly scoped, revocable token that you are comfortable sending to api.talesofai.com, avoid sensitive personal or confidential prompts/reference IDs, and prefer a pinned or checksum-verified install source.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T08 · Insecure Dependencies

Warning
Location
README.md:9
Finding

Unpinned Remote Installation Chain

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Note
Location
caricatureportraitgenerator.js:2
Finding

API Token Exposed Through Command-Line Arguments

Content
View full analysis
"your prompt" --token YOUR_TOKEN ``` The README also recommends: ```bash node caricatureportraitgenerator.js "middle-aged professor with wild white hair and oversized glasses, academic caricature style" --token "$NETA_TOKEN" ``` Although shell expansion obtains the value from an environment variable in the latter example, the expanded token is still passed as a process argument. ### Technical Analysis Command-line arguments are not an appropriate primary channel for sensitive credentials. Depending on the operating system and execution environment, process arguments may be visible through process inspection interfaces, monitoring systems, diagnostic reports, audit tooling, or job metadata while the process is running. A literal invocation such as `--token YOUR_TOKEN` may additionally preserve the credential in shell history. The implementation reads the token directly from `process.argv` and provides no safer input mechanism. The token is subsequently sent in the `x-token` HTTPS header to the declared API endpoint. That transmission is consistent with the Skill's stated image-generation function; the issue is the local credential ingestion method, not the intended authenticated HTTPS request. ### Attack Path 1. A user follows the documentation and supplies a valid Neta API token through `--token`. 2. The token becomes part of the pr ...[truncated 908 chars]
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The README instructs users to send arbitrary prompts and a bearer API token to a third-party image-generation service, but it does not disclose privacy, retention, or data-sharing implications. Users may unknowingly submit sensitive personal descriptions, likeness data, or other confidential content to an external provider, increasing the risk of data exposure or misuse.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

The install instruction uses npx skills add omactiengartelle/caricature-portrait-generator without pinning an exact version, so users may fetch whatever package version is current at execution time. This creates a supply-chain risk: a compromised upstream package, account takeover, or malicious later update could result in unreviewed code being installed or executed.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The script explicitly instructs the user to obtain a token from Neta, then forwards that token in the x-token header to a different host, api.talesofai.com. That mismatch is dangerous because it can trick users into surrendering credentials under false pretenses, and the token may grant access or be replayed against systems the user did not intend to trust.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill metadata says it uses the Neta AI API, but the code actually sends prompts and token-authenticated requests to api.talesofai.com. This creates a deceptive trust boundary: users may provide a token obtained for one service while their data and credentials are transmitted to a different backend, enabling credential misuse, data exfiltration, or unauthorized billing/abuse.

Content

No source excerpt is available for this finding.