Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward AIsa Perplexity API helper, with the main risk being that user queries and optional system prompts are sent to a third-party service.
Install this only if you are comfortable sending the query text and any optional system instruction to AIsa's Perplexity API using your AISA_API_KEY. Avoid submitting secrets, regulated data, private documents, or confidential internal context unless that sharing is approved for your environment.
The skill clearly requires environment access (AISA_API_KEY) and performs outbound network requests, but it does not declare an explicit tool scope such as permissions or allowed-tools. That omission weakens least-privilege controls and makes it harder for users or the platform to understand and constrain what the skill can access and transmit.
The skill description says it is for Perplexity-powered answers but does not warn that user prompts, queries, and optional system messages are sent to an external API service. This can cause unintentional disclosure of sensitive user data or internal context to a third party because users are not clearly informed at the point of use.
This section introduces curl-based use of an external API and demonstrates sending user-provided content off-platform. In this skill context, external transmission is expected functionality, but it is still security-relevant because prompts and system instructions may contain sensitive data and are sent to a third party.
--system "Respond in markdown with an executive summary first."
## Curl Examples
### Sonar
The Sonar example posts messages and an API bearer token to https://api.aisa.one, confirming direct external transmission of user content to a third-party service. The behavior is aligned with the skill's purpose, but it becomes dangerous if users assume local processing or include confidential information in the query.
curl -X POST "https://api.aisa.one/apis/v1/perplexity/sonar" \
-H "Authorization: Bearer $AISA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
The Sonar Pro example sends user messages to the same external API endpoint family, creating the same data-exposure risk as the basic Sonar example. Because this model is positioned for stronger synthesis and comparison, users may be more likely to submit richer internal context, increasing potential disclosure impact.
curl -X POST "https://api.aisa.one/apis/v1/perplexity/sonar-pro" \
-H "Authorization: Bearer $AISA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
The Sonar Reasoning Pro example transmits prompts externally for multi-step reasoning, which can include detailed proprietary or strategic information. The skill does not provide compensating warnings or guardrails, so users may unknowingly export sensitive reasoning tasks to an external provider.
curl -X POST "https://api.aisa.one/apis/v1/perplexity/sonar-reasoning-pro" \
-H "Authorization: Bearer $AISA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
The Sonar Deep Research example sends potentially large, detailed research requests to an external API, increasing the chance of transmitting extensive sensitive context. In a deep-research workflow, the volume and specificity of submitted material can magnify confidentiality and compliance risk if users are not warned.
curl -X POST "https://api.aisa.one/apis/v1/perplexity/sonar-deep-research" \
-H "Authorization: Bearer $AISA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
class PerplexitySearchClient:
"""Minimal client for AIsa Perplexity Sonar endpoints."""
BASE_URL = "https://api.aisa.one/apis/v1"
def __init__(self, api_key: Optional[str] = None):
self.api_key = api_key or os.environ.get("AISA_API_KEY")
This code builds a request containing the user-provided query and optional system instruction, then sends it via HTTP POST to an external API. While the file's docstring describes usage, it does not clearly warn users that their input will be transmitted to a remote service.
The client reads the AISA_API_KEY environment variable, which is a sensitive credential source. Although this is typical for API clients, the file does not include a warning or note informing users that the skill depends on a credential from the environment.
No suspicious patterns detected.