Skip to main content

Overview

When a secret is detected, Cysmiq can verify whether it is still active by making API calls to the provider. This page lists supported verification providers, the key names we verify, and any extra materials needed for verification.

Supported providers

Cloud & Infrastructure

ProviderKey nameExtra materials
AWSSecret access keyAccess key ID
GCPService account key (JSON)-
DigitalOceanAPI key-
CloudflareAPI token-
Cloudflare CAOrigin CA key-
Cloudflare GlobalGlobal API keyEmail
VercelAccess token-
NetlifyAPI key-
HerokuAPI key-
DatabricksAPI keyInstance name
InfuraAPI key-

AI & Machine Learning

ProviderKey nameExtra materials
OpenAIAPI key-
AnthropicAPI key-

Payment & Commerce

ProviderKey nameExtra materials
StripeAPI key-
SquareApplication access token-
ShopifyTokenDomain
ChargebeeAPI keySite ID

Communication & Email

ProviderKey nameExtra materials
TwilioTokenAccount SID
SendGridAPI key-
MailgunPrimary API key-
MailchimpAPI key-
SlackWebhook-
TelegramBot API key-
Microsoft TeamsWebhook-

Developer Tools & CI/CD

ProviderKey nameExtra materials
CysmiqAPI key-
GitHubPersonal access token (PAT), Classic token-
CircleCIToken-
Travis CIAPI key-
SonarCloudAPI key-
CodeClimateKey-
StackHawkAPI key-
PostmanAPI key-
ArtifactoryTokenJFrog URL

Package Registries

ProviderKey nameExtra materials
npmToken (.npmrc), Token-
NuGetAPI key-
RubyGemsAPI key-

Data & Analytics

ProviderKey nameExtra materials
CloudinaryAPI secretCloud name, API key
AirtablePAT-
PolygonAPI key-
ShodanAPI key-
AbuseIPDBAPI key-

Other Services

ProviderKey nameExtra materials
AgoraCustomer secretCustomer ID
AlchemyAPI key-
BeamerAPI key-
BitkubAPI secretAPI key
DropboxAPI key-
FastlyAPI key-
Google APIsAPI key-
LinearAPI key-
AdafruitAPI key-
AbyssaleAPI key-
SurgeAIAPI key-

Extra materials

Some providers need additional context beyond the secret itself to perform verification. When extra materials are missing, verification may be skipped.

LLM-based validation

Secrets without API verification support are validated using LLM-based analysis. This includes:
  • High-entropy tokens in configuration files
  • Generic API key patterns
  • Private keys and certificates
  • Provider-ambiguous credentials
LLM validation evaluates the detection context to reduce false positives.