PraisonAI before 4.6.78 exposes the MCP HTTP-stream transport without authentication by default: the CLI --api-key option defaults to None, and the server only enforces Authorization/Bearer checks when an API key is configured. When an operator runs 'praisonai mcp serve --transport http-stream' without an API key, an unauthenticated client (no Authorization header, and no Origin header, which is also permitted) can initialize a session, enumerate the available tools (tools/list), and invoke tools (tools/call). Additionally, the dispatcher forwards tool-call arguments to handlers without validating them against the advertised inputSchema. The server binds to 127.0.0.1 by default, so remote exploitation requires the operator to bind to a network-accessible address (e.g., --host 0.0.0.0).
Casky was already ahead
This CVE exploits attack patterns that Casky's 322matched skills already investigate — long before this vulnerability was disclosed. Claude's reasoning model maps these techniques to MITRE ATT&CK, so practitioners who ran these skills have already seen the threat behaviour in their findings.
PraisonAI versions before 4.6.78 contain a critical authentication bypass in the Model Context Protocol (MCP) HTTP-stream transport. The vulnerability exists because the CLI defaults the --api-key option to None, and the server only validates Authorization Bearer tokens when an API key is explicitly configured. When operators deploy 'praisonai mcp serve --transport http-stream' without specifying an API key, unauthenticated clients can connect without credentials, enumerate available tools via the tools/list endpoint, and execute arbitrary tool invocations. This affects any organization running vulnerable PraisonAI instances in networked environments, particularly those exposing the MCP server to untrusted networks or within shared infrastructure where lateral movement is possible. The vulnerability's CVSS 7.3 rating reflects the high impact of unauthenticated access to potentially sensitive tools and data processing capabilities.
Casky's 322 mapped skills leverage Claude's extended reasoning to detect the attack patterns underlying this CVE by correlating techniques from both TA0001 (Reconnaissance) and TA0043 (Resource Development). Practitioners using Casky would identify reconnaissance activity through CWE-20 (improper input validation) detection patterns: unauthenticated HTTP requests to MCP endpoints lacking proper authorization headers, followed by tools/list enumeration queries. The skills would flag tool invocation attempts from clients without valid Bearer tokens, correlating these with MITRE technique patterns for initial access and command & control communication. Security teams would see findings highlighting the dangerous default configuration, the absence of authentication checks in unauthenticated paths, and the exposure of administrative/tool interfaces—allowing them to prioritize patching, enforce network segmentation, and implement compensating controls like reverse proxy authentication before upgrading to 4.6.78 or later.
Composite risk scoring from EPSS, CISA KEV, Shodan, and GreyNoise — 21 security APIs correlated into a single Casky Risk Score. Coming in Casky Pro. Join early access →
These skills use Claude AI's reasoning model to surface findings in the same attack categories as CVE-2026-61427.
Casky has 322 skills that investigate the attack patterns behind CVE-2026-61427. Run one and get CVSS-scored findings in 3 minutes.
Run the skill that detects this →Access with Stolen Session Cookie
penetration testing · medium
Account Access Removal
cloud security · low
Account Manipulation
cloud security · low
Account Manipulation: Account Linking
cloud security · low
Account Manipulation: Change Account Details
cloud security · low
Account Manipulation: Change of Payment Details
phishing defense · medium
Account Takeover
red teaming · high
Account Takeover
phishing defense · medium
Account Takeover
red teaming · high
Account Takeover: Exposed Login Credential
red teaming · high
Account Takeover: Exposed Login Credential
phishing defense · medium
achieving-cmmc-level-2-compliance
compliance governance · low
analyzing-apt-group-with-mitre-navigator
threat intelligence · low
analyzing-campaign-attribution-evidence
threat intelligence · low
analyzing-cloud-storage-access-patterns
cloud security · low
analyzing-cyber-kill-chain
threat intelligence · low
analyzing-ios-app-security-with-objection
mobile security · low
analyzing-malicious-url-with-urlscan
phishing defense · medium
analyzing-malware-family-relationships-with-malpedia
threat intelligence · low
analyzing-office365-audit-logs-for-compromise
cloud security · low
analyzing-threat-actor-ttps-with-mitre-attack
threat intelligence · low
analyzing-threat-actor-ttps-with-mitre-navigator
threat intelligence · low
analyzing-threat-intelligence-feeds
threat intelligence · low
analyzing-threat-landscape-with-misp
threat intelligence · low
auditing-aws-s3-bucket-permissions
cloud security · low
auditing-azure-active-directory-configuration
cloud security · low
auditing-cloud-with-cis-benchmarks
cloud security · low
auditing-gcp-iam-permissions
cloud security · low
auditing-terraform-infrastructure-for-security
cloud security · low
auditing-tls-certificate-transparency-logs
threat intelligence · low
automating-ioc-enrichment
threat intelligence · low
Browser Session Hijacking
cloud security · low
Brute Force: Credential Stuffing
threat intelligence · low
building-adversary-infrastructure-tracking-system
threat intelligence · low
building-attack-pattern-library-from-cti-reports
threat intelligence · low
building-c2-infrastructure-with-sliver-framework
red teaming · high
building-c2-redirector-infrastructure
red teaming · high
building-cloud-siem-with-sentinel
cloud security · low
building-devsecops-pipeline-with-gitlab-ci
devsecops · low
building-ioc-defanging-and-sharing-pipeline
threat intelligence · low
building-ioc-enrichment-pipeline-with-opencti
threat intelligence · low
building-patch-tuesday-response-process
vulnerability management · medium
building-red-team-c2-infrastructure-with-havoc
red teaming · high
building-threat-actor-profile-from-osint
threat intelligence · low
building-threat-feed-aggregation-with-misp
threat intelligence · low
building-threat-intelligence-platform
threat intelligence · low
building-vulnerability-aging-and-sla-tracking
vulnerability management · medium
building-vulnerability-dashboard-with-defectdojo
vulnerability management · medium
building-vulnerability-exception-tracking-system
vulnerability management · medium
bypassing-authentication-with-forced-browsing
web application security · medium
coercing-authentication-with-coercer-petitpotam
red teaming · high
collecting-open-source-intelligence
threat intelligence · low
collecting-threat-intelligence-with-misp
threat intelligence · low
conducting-api-security-testing
penetration testing · medium
conducting-cloud-penetration-testing
cloud security · low
conducting-cyber-risk-assessment-with-nist-800-30
compliance governance · low
conducting-domain-persistence-with-dcsync
red teaming · high
conducting-external-reconnaissance-with-osint
penetration testing · medium
conducting-full-scope-red-team-engagement
red teaming · high
conducting-internal-network-penetration-test
penetration testing · medium
conducting-internal-reconnaissance-with-bloodhound-ce
red teaming · high
conducting-mobile-app-penetration-test
penetration testing · medium
conducting-network-penetration-test
penetration testing · medium
conducting-pass-the-ticket-attack
red teaming · high
conducting-wireless-network-penetration-test
penetration testing · medium
Convert to Cryptocurrency
cloud security · low
correlating-threat-campaigns
threat intelligence · low
Create Fake Materials: Fake Website
phishing defense · medium
Create Fake Materials: Fake Website
threat intelligence · low
Create Fake Materials: Fake Website
penetration testing · medium
Create Fake Materials: Fake Website
phishing defense · medium
Create Fake Materials: Fake Website
phishing defense · medium
Create Fake Materials: Fake Website
phishing defense · medium
Delete Relevant Emails
phishing defense · medium
detecting-api-enumeration-attacks
api security · medium
detecting-aws-guardduty-findings-automation
cloud security · low
detecting-aws-iam-privilege-escalation
cloud security · low
detecting-azure-lateral-movement
cloud security · low
detecting-azure-service-principal-abuse
cloud security · low
detecting-azure-storage-account-misconfigurations
cloud security · low
detecting-broken-object-property-level-authorization
api security · medium
detecting-cloud-threats-with-guardduty
cloud security · low
detecting-misconfigured-azure-storage
cloud security · low
detecting-s3-data-exfiltration-attempts
cloud security · low
detecting-serverless-function-injection
cloud security · low
detecting-shadow-api-endpoints
api security · medium
detecting-shadow-it-cloud-usage
cloud security · low
detecting-suspicious-oauth-application-consent
cloud security · low
Electronic Funds Transfer: Wire Transfer
phishing defense · medium
Electronic Funds Transfer: Wire Transfer
threat intelligence · low
Email Spoofing
threat intelligence · low
emulating-cloud-attacks-with-stratus-red-team
cloud security · low
enumerating-cloud-with-cloudfox
cloud security · low
evaluating-threat-intelligence-platforms
threat intelligence · low
executing-active-directory-attack-simulation
penetration testing · medium
executing-nist-rmf-authorization-to-operate
compliance governance · low
executing-red-team-engagement-planning
red teaming · high
executing-red-team-exercise
penetration testing · medium
exploiting-active-directory-certificate-services-esc1
red teaming · high
exploiting-active-directory-with-bloodhound
red teaming · high
exploiting-adcs-with-certipy
red teaming · high
exploiting-api-injection-vulnerabilities
api security · medium
exploiting-aws-with-pacu
cloud security · low
exploiting-broken-function-level-authorization
api security · medium
exploiting-broken-link-hijacking
web application security · medium
exploiting-constrained-delegation-abuse
red teaming · high
exploiting-deeplink-vulnerabilities
mobile security · low
exploiting-excessive-data-exposure-in-api
api security · medium
exploiting-http-request-smuggling
web application security · medium
exploiting-idor-vulnerabilities
web application security · medium
exploiting-insecure-data-storage-in-mobile
mobile security · low
exploiting-insecure-deserialization
web application security · medium
exploiting-jwt-algorithm-confusion-attack
api security · medium
exploiting-kerberoasting-with-impacket
red teaming · high
exploiting-mass-assignment-in-rest-apis
web application security · medium
exploiting-ms17-010-eternalblue-vulnerability
red teaming · high
exploiting-nopac-cve-2021-42278-42287
red teaming · high
exploiting-nosql-injection-vulnerabilities
web application security · medium
exploiting-oauth-misconfiguration
web application security · medium
exploiting-prototype-pollution-in-javascript
web application security · medium
exploiting-race-condition-vulnerabilities
web application security · medium
exploiting-server-side-request-forgery
web application security · medium
exploiting-sql-injection-vulnerabilities
penetration testing · medium
exploiting-sql-injection-with-sqlmap
web application security · medium
exploiting-template-injection-vulnerabilities
web application security · medium
exploiting-type-juggling-vulnerabilities
web application security · medium
exploiting-vulnerabilities-with-metasploit-framework
vulnerability management · medium
exploiting-websocket-vulnerabilities
web application security · medium
exploiting-zerologon-vulnerability-cve-2020-1472
red teaming · high
Gather Customer Information
threat intelligence · low
generating-threat-intelligence-reports
threat intelligence · low
hunting-advanced-persistent-threats
threat intelligence · low
Impersonate Account Holder
phishing defense · medium
Impersonate Account Holder
phishing defense · medium
implementing-api-abuse-detection-with-rate-limiting
api security · medium
implementing-api-gateway-security-controls
api security · medium
implementing-api-key-security-controls
api security · medium
implementing-api-rate-limiting-and-throttling
api security · medium
implementing-api-schema-validation-security
api security · medium
implementing-api-security-posture-management
api security · medium
implementing-api-security-testing-with-42crunch
api security · medium
implementing-api-threat-protection-with-apigee
api security · medium
implementing-aqua-security-for-container-scanning
devsecops · low
implementing-attack-path-analysis-with-xm-cyber
vulnerability management · medium
implementing-aws-config-rules-for-compliance
cloud security · low
implementing-aws-macie-for-data-classification
cloud security · low
implementing-aws-nitro-enclave-security
cloud security · low
implementing-aws-security-hub
cloud security · low
implementing-aws-security-hub-compliance
cloud security · low
implementing-azure-defender-for-cloud
cloud security · low
implementing-cloud-dlp-for-data-protection
cloud security · low
implementing-cloud-security-posture-management
cloud security · low
implementing-cloud-trail-log-analysis
cloud security · low
implementing-cloud-vulnerability-posture-management
vulnerability management · medium
implementing-cloud-waf-rules
cloud security · low
implementing-cloud-workload-protection
cloud security · low
implementing-code-signing-for-artifacts
devsecops · low
implementing-continuous-security-validation-with-bas
vulnerability management · medium
implementing-diamond-model-analysis
threat intelligence · low
implementing-dmarc-dkim-spf-email-security
phishing defense · medium
implementing-email-sandboxing-with-proofpoint
phishing defense · medium
implementing-epss-score-for-vulnerability-prioritization
vulnerability management · medium
implementing-fuzz-testing-in-cicd-with-aflplusplus
devsecops · low
implementing-gcp-binary-authorization
cloud security · low
implementing-gcp-organization-policy-constraints
cloud security · low
implementing-gcp-vpc-firewall-rules
cloud security · low
implementing-gdpr-data-protection-controls
compliance governance · low
implementing-github-advanced-security-for-code-scanning
devsecops · low
implementing-hipaa-security-rule-safeguards
compliance governance · low
implementing-infrastructure-as-code-security-scanning
devsecops · low
implementing-iso-27001-information-security-management
compliance governance · low
implementing-mobile-application-management
mobile security · low
implementing-patch-management-workflow
vulnerability management · medium
implementing-pci-dss-compliance-controls
compliance governance · low
implementing-policy-as-code-with-open-policy-agent
devsecops · low
implementing-rapid7-insightvm-for-scanning
vulnerability management · medium
implementing-secret-scanning-with-gitleaks
devsecops · low
implementing-secrets-management-with-vault
cloud security · low
implementing-secrets-scanning-in-ci-cd
devsecops · low
implementing-security-information-sharing-with-stix2
threat intelligence · low
implementing-semgrep-for-custom-sast-rules
devsecops · low
implementing-stix-taxii-feed-integration
threat intelligence · low
implementing-taxii-server-with-opentaxii
threat intelligence · low
implementing-threat-intelligence-lifecycle-management
threat intelligence · low
implementing-vulnerability-management-with-greenbone
vulnerability management · medium
implementing-vulnerability-remediation-sla
vulnerability management · medium
implementing-vulnerability-sla-breach-alerting
vulnerability management · medium
implementing-web-application-logging-with-modsecurity
web application security · medium
implementing-zero-trust-in-cloud
cloud security · low
implementing-zero-trust-network-access
cloud security · low
integrating-dast-with-owasp-zap-in-pipeline
devsecops · low
integrating-sast-into-github-actions-pipeline
devsecops · low
intercepting-mobile-traffic-with-burpsuite
mobile security · low
managing-intelligence-lifecycle
threat intelligence · low
managing-third-party-vendor-risk
compliance governance · low
mapping-attack-paths-with-bloodhound-ce
red teaming · high
mapping-mitre-attack-techniques
threat intelligence · low
modeling-threats-with-opencti
threat intelligence · low
moving-laterally-with-netexec
penetration testing · medium
operating-havoc-c2
red teaming · high
operating-sliver-c2
red teaming · high
operationalizing-misp-threat-feeds
threat intelligence · low
performing-active-directory-bloodhound-analysis
red teaming · high
performing-active-directory-penetration-test
penetration testing · medium
performing-active-directory-vulnerability-assessment
vulnerability management · medium
performing-agentless-vulnerability-scanning
vulnerability management · medium
performing-ai-driven-osint-correlation
threat intelligence · low
performing-android-app-static-analysis-with-mobsf
mobile security · low
performing-api-fuzzing-with-restler
api security · medium
performing-api-inventory-and-discovery
api security · medium
performing-api-rate-limiting-bypass
api security · medium
performing-api-security-testing-with-postman
api security · medium
performing-asset-criticality-scoring-for-vulns
vulnerability management · medium
performing-authenticated-scan-with-openvas
vulnerability management · medium
performing-authenticated-vulnerability-scan
vulnerability management · medium
performing-aws-account-enumeration-with-scout-suite
cloud security · low
performing-aws-privilege-escalation-assessment
cloud security · low
performing-blind-ssrf-exploitation
web application security · medium
performing-clickjacking-attack-test
web application security · medium
performing-cloud-asset-inventory-with-cartography
cloud security · low
performing-cloud-forensics-with-aws-cloudtrail
cloud security · low
performing-cloud-log-forensics-with-athena
cloud security · low
performing-cloud-native-forensics-with-falco
cloud security · low
performing-cloud-native-threat-hunting-with-aws-detective
cloud security · low
performing-cloud-penetration-testing-with-pacu
cloud security · low
performing-container-image-hardening
devsecops · low
performing-content-security-policy-bypass
web application security · medium
performing-csrf-attack-simulation
web application security · medium
performing-cve-prioritization-with-kev-catalog
vulnerability management · medium
performing-dark-web-monitoring-for-threats
threat intelligence · low
performing-directory-traversal-testing
web application security · medium
performing-dmarc-policy-enforcement-rollout
phishing defense · medium
performing-dynamic-analysis-of-android-app
mobile security · low
performing-external-network-penetration-test
penetration testing · medium
performing-gcp-penetration-testing-with-gcpbucketbrute
cloud security · low
performing-gcp-security-assessment-with-forseti
cloud security · low
performing-graphql-depth-limit-attack
api security · medium
performing-graphql-introspection-attack
api security · medium
performing-graphql-security-assessment
web application security · medium
performing-http-parameter-pollution-attack
web application security · medium
performing-indicator-lifecycle-management
threat intelligence · low
performing-ios-app-security-assessment
mobile security · low
performing-iot-security-assessment
penetration testing · medium
performing-ip-reputation-analysis-with-shodan
threat intelligence · low
performing-jwt-none-algorithm-attack
api security · medium
performing-kerberoasting-attack
red teaming · high
performing-lateral-movement-with-wmiexec
red teaming · high
performing-malware-hash-enrichment-with-virustotal
threat intelligence · low
performing-malware-ioc-extraction
threat intelligence · low
performing-mobile-app-certificate-pinning-bypass
mobile security · low
performing-nist-csf-maturity-assessment
compliance governance · low
performing-open-source-intelligence-gathering
red teaming · high
performing-osint-with-spiderfoot
threat intelligence · low
performing-physical-intrusion-assessment
red teaming · high
performing-privilege-escalation-assessment
penetration testing · medium
performing-privilege-escalation-on-linux
red teaming · high
performing-sca-dependency-scanning-with-snyk
devsecops · low
performing-second-order-sql-injection
web application security · medium
performing-security-headers-audit
web application security · medium
performing-serverless-function-security-review
cloud security · low
performing-soap-web-service-security-testing
api security · medium
performing-subdomain-enumeration-with-subfinder
web application security · medium
performing-thick-client-application-penetration-test
penetration testing · medium
performing-threat-emulation-with-atomic-red-team
threat intelligence · low
performing-threat-intelligence-sharing-with-misp
threat intelligence · low
performing-threat-landscape-assessment-for-sector
threat intelligence · low
performing-threat-modeling-with-owasp-threat-dragon
devsecops · low
performing-vulnerability-scanning-with-nessus
penetration testing · medium
performing-web-application-firewall-bypass
web application security · medium
performing-web-application-penetration-test
penetration testing · medium
performing-web-application-scanning-with-nikto
vulnerability management · medium
performing-web-application-vulnerability-triage
vulnerability management · medium
performing-web-cache-deception-attack
web application security · medium
performing-web-cache-poisoning-attack
web application security · medium
performing-wireless-network-penetration-test
penetration testing · medium
Phishing
threat intelligence · low
Phishing
threat intelligence · low
Phone Number Spoofing: Official Phone Number Spoofing
red teaming · high
prioritizing-vulnerabilities-with-cvss-scoring
vulnerability management · medium
processing-stix-taxii-feeds
threat intelligence · low
profiling-threat-actor-groups
threat intelligence · low
relaying-ntlm-for-adcs-esc8
red teaming · high
remediating-s3-bucket-misconfiguration
cloud security · low
reverse-engineering-ios-app-with-frida
mobile security · low
scanning-containers-with-trivy-in-cicd
devsecops · low
scanning-iac-and-images-with-trivy
devsecops · low
scanning-infrastructure-with-nessus
vulnerability management · medium
securing-api-gateway-with-aws-waf
cloud security · low
securing-aws-lambda-execution-roles
cloud security · low
securing-azure-with-microsoft-defender
cloud security · low
securing-container-registry-images
cloud security · low
securing-github-actions-workflows
devsecops · low
securing-kubernetes-on-cloud
cloud security · low
securing-serverless-functions
cloud security · low
Stage Capabilities: SEO Poisoning
threat intelligence · low
testing-android-intents-for-vulnerabilities
mobile security · low
testing-api-authentication-weaknesses
api security · medium
testing-api-for-broken-object-level-authorization
api security · medium
testing-api-for-mass-assignment-vulnerability
api security · medium
testing-api-security-with-owasp-top-10
web application security · medium
testing-cors-misconfiguration
web application security · medium
testing-for-broken-access-control
web application security · medium
testing-for-business-logic-vulnerabilities
web application security · medium
testing-for-email-header-injection
web application security · medium
testing-for-host-header-injection
web application security · medium
testing-for-json-web-token-vulnerabilities
web application security · medium
testing-for-open-redirect-vulnerabilities
web application security · medium
testing-for-sensitive-data-exposure
web application security · medium
testing-for-xml-injection-vulnerabilities
web application security · medium
testing-for-xss-vulnerabilities
penetration testing · medium
testing-for-xss-vulnerabilities-with-burpsuite
web application security · medium
testing-for-xxe-injection-vulnerabilities
web application security · medium
testing-jwt-token-security
web application security · medium
testing-mobile-api-authentication
mobile security · low
testing-oauth2-implementation-flaws
api security · medium
testing-websocket-api-security
api security · medium
triaging-vulnerabilities-with-ssvc-framework
vulnerability management · medium
Use Alternate Authentication Material: Application Access Token
cloud security · low
Use Alternate Authentication Material: Application Access Token
cloud security · low
© 2026 Casky.AI, Inc. · AI Security Investigation