ClawHub
Back to skill

Security audit

Perplexity Search Skill

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Perplexity web-search skill that sends user search queries to Perplexity using a user-provided API key, with no evidence of hidden persistence, local file access, or unrelated behavior.

Install this only if you are comfortable with web-search prompts being sent to Perplexity and with usage being billed to your Perplexity API key. Avoid using it for prompts containing secrets or private local data unless you intentionally want that text included in the external search request.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (10)

Credential Access

High
Category
Privilege Escalation
Content
Set your Perplexity API key in the environment before starting OpenClaw:

```bash
# Add to ~/.openclaw/.env
PERPLEXITY_API_KEY=pplx-your-key-here
```
Confidence
60% confidence
Finding
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The README states the skill is auto-invoked on web-search intent with no visible narrowing criteria, which can cause the agent to send user prompts to an external service more often than users expect. Overly broad invocation increases the chance of transmitting sensitive or irrelevant prompt content to a third party and can create unintended tool use.

Vague Triggers

Medium
Confidence
95% confidence
Finding
Claiming activation for 'anything web-related' is excessively broad and ambiguous, encouraging a tool-routing policy that may trigger on many ordinary prompts. In practice, this can expand external data sharing beyond user intent and increase privacy, compliance, and prompt-injection exposure through unnecessary browsing.

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill declares capabilities requiring environment-variable access and network egress, but does not explicitly scope or declare permissions. That omission increases the chance of unintended execution or overbroad access because reviewers and runtime policy engines cannot easily enforce least privilege from the manifest alone.

Vague Triggers

Medium
Confidence
97% confidence
Finding
The trigger phrases are broad everyday terms like 'current', 'recent', and 'find out', which can cause the skill to auto-invoke in contexts where the user did not intend a web search. Because this skill performs network access and consumes an API-backed credential, accidental invocation can leak user queries externally and incur unnecessary cost.

External Transmission

Medium
Category
Data Exfiltration
Content
# ── Constants ──────────────────────────────────────────────────────────────────

API_URL = "https://api.perplexity.ai/chat/completions"
DEFAULT_MODEL = "sonar"

ALLOWED_MODELS = {
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
# ── Constants ──────────────────────────────────────────────────────────────────

API_URL = "https://api.perplexity.ai/chat/completions"
DEFAULT_MODEL = "sonar"

ALLOWED_MODELS = {
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
# ── Constants ──────────────────────────────────────────────────────────────────

API_URL = "https://api.perplexity.ai/chat/completions"
DEFAULT_MODEL = "sonar"

ALLOWED_MODELS = {
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
# ── Constants ──────────────────────────────────────────────────────────────────

API_URL = "https://api.perplexity.ai/chat/completions"
DEFAULT_MODEL = "sonar"

ALLOWED_MODELS = {
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

External Transmission

Medium
Category
Data Exfiltration
Content
# ── Constants ──────────────────────────────────────────────────────────────────

API_URL = "https://api.perplexity.ai/chat/completions"
DEFAULT_MODEL = "sonar"

ALLOWED_MODELS = {
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Static analysis

No suspicious patterns detected.