Security audit
Security checks for vulnerabilities and agentic risk
This is a narrow Perplexity/AIsa search skill that sends user queries to AIsa using an API key, which is disclosed and matches its stated purpose.
Install this only if you intend to use AIsa/Perplexity Sonar from your agent. Do not submit secrets, private documents, regulated data, or internal prompts unless you are comfortable sending that content to api.aisa.one under your AIsa account.
The skill explicitly requires environment access to AISA_API_KEY and performs outbound network requests, but it does not declare a restrictive tool scope such as permissions or allowed-tools. In an agent ecosystem, missing scope declarations weakens policy enforcement and can let the skill run with broader capabilities than intended.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
--system "Respond in markdown with an executive summary first."
## Curl Examples
### Sonar
The skill sends user-supplied content to an external service at api.aisa.one using a bearer token from the environment. While this is the core function of the skill, it is still a real data-exfiltration boundary because prompts, system instructions, and potentially sensitive user data are transmitted off-platform.
curl -X POST "https://api.aisa.one/apis/v1/perplexity/sonar" \
-H "Authorization: Bearer $AISA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
This example shows another external POST to the same third-party API with user content and authorization headers. Even though it is functionally intended, the transmission creates confidentiality and compliance risk if an agent forwards sensitive prompts or embedded data without strong disclosure and scope controls.
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 reasoning-pro endpoint similarly transmits prompt content to an external API, increasing the chance that lengthy analytical prompts contain proprietary or sensitive information. The skill context makes the behavior expected, but not harmless, because the data flow crosses a trust boundary.
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 deep-research endpoint is especially sensitive because it encourages long-form research requests, which may include larger volumes of proprietary context or internal documents in the prompt. That makes this external transmission more dangerous than a short query, despite being consistent with the skill's intended purpose.
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 packages the user's query and optional system instruction into a request body and sends it to an external HTTPS endpoint. Although the file is a network client by purpose, there is no explicit disclosure in code comments, help text, or runtime output that user-provided content will be transmitted to a remote service.
No suspicious patterns detected.