Back to skill

Security audit

custom-infographic

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because it found an unsafe boundary when processing images returned by an external service and privacy and scope concerns; this is a review warning, not a finding of malicious intent.

Consider the privacy of any source material before using the skill, since scanners report that source-derived content may reach external services. Review the broad activation trigger and the image-processing warning before installing. Scanning was partial: although all listed components were scanned, some local references were unresolved, so the evidence is incomplete. A separate scanner reported no findings, but that does not establish that the skill is risk-free.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (7)

Tainted flow: 'req' from os.environ.get (line 49, credential/environment) → urllib.request.urlopen (network output)

Critical
Category
Data Flow
Confidence
90% confidence
Finding

Credentials or environment variables flow to a network sink. This is a high-confidence indicator of credential exfiltration.

Content

Scanner excerpt · scripts/generate_image.py (reported line 61)May include surrounding context.

python
)

    try:
        with urllib.request.urlopen(req) as resp:
            data = json.loads(resp.read().decode("utf-8"))
    except urllib.error.HTTPError as e:
        body = e.read().decode("utf-8", errors="replace")

Unvalidated Output Injection

High
Category
Output Handling
Confidence
95% confidence
Finding

Bytes returned through the external service are written without format or size validation and then passed directly to ImageMagick. A malicious or compromised response could exercise unexpected decoders or cause excessive memory and disk consumption; this is an unsafe image-processing boundary, not demonstrated shell injection.

Content

Scanner excerpt · scripts/generate_image.py (reported line 109)May include surrounding context.

python
# image is already on disk and usable.
    tmp = args.output + ".opt"
    try:
        subprocess.run(
            ["magick", args.output, "-strip", "-colors", "256", tmp],
            stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL,
            check=True, timeout=180,

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, network, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The trigger “可视化” (visualization) is broader than infographic creation and has no exclusion conditions.

Medium
Category
Trigger Abuse
Confidence
95% confidence
Finding

The skill explicitly activates on a general request for visualization, which could instead mean a chart, interactive dashboard, or data-exploration view. No exclusion or clarification requirement distinguishes those requests from infographic generation.

Content

No source excerpt is available for this finding.

The image-generation workflow lacks an explicit privacy warning about sending source-derived content to external model services.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

The workflow sends source-derived text to OpenRouter and a downstream model provider without explicitly confirming that external disclosure is authorized. Secret stripping helps, but does not protect confidential business information, personal data, or other sensitive document content; design-option confirmation is not a privacy check.

Content

No source excerpt is available for this finding.

External image-generation handoff lacks a warning about disclosing source-document content.

Medium
Category
Output Handling
Confidence
97% confidence
Finding

The example recommends pasting a prompt containing verbatim source-derived material into external services without checking whether the document may be shared. Manual handoff reduces automatic disclosure risk, but users may still expose confidential or personal information by following this guidance.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/generate_image.py (reported line 109)May include surrounding context.

python
# image is already on disk and usable.
    tmp = args.output + ".opt"
    try:
        subprocess.run(
            ["magick", args.output, "-strip", "-colors", "256", tmp],
            stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL,
            check=True, timeout=180,