PentestMate

PivotGG is an AI-powered platform designed to streamline incident response and threat investigation. Key features include:AI-driven pivot workflows: Automatically generate investigation paths based on IOCs.AI-generated searches: Create Splunk/KQL searches for rapid detection.Detection opportunities: Identify potential threats and vulnerabilities.Ideal for:Security analystsThreat huntersIncident respondersPivotGG empowers security professionals to investigate threats faster and more effectively, saving valuable time and resources.

Proxed.AI is a security and management platform for AI APIs. It allows developers to secure API keys, verify device authenticity, and structure AI outputs without needing a backend. Key features include:API Key Protection: Securely lock down your AI API credentials.DeviceCheck Integration: Verify device authenticity using Apple's DeviceCheck.Structured AI Responses: Structure and manage AI outputs for better data handling.Open Source & Self-Hosting: Offers flexibility with open-source code and self-hosting options.Ideal for:iOS Developers integrating AI APIsStartups looking for cost-effective AI security solutionsDevelopers needing structured data from AI modelsProxed.AI simplifies AI API security and management, offering a no-backend solution with open-source flexibility and cost-saving features.

AgentFacts is an open-source SDK for verifiable AI agent metadata.Generate signed profiles with DIDs, capabilities, and provenance for LangChain, Hugging Face, and custom agents. All self-hosted, framework-agnostic, and tamper-proof.

SignalVault is an infrastructure layer that sits between your application and AI providers like OpenAI and Anthropic. Every prompt and response flows through SignalVault's proxy, where it's encrypted, scanned for violations, and logged with a complete audit trail. THE PROBLEMTeams ship AI features fast. Audit logs come later — if at all. PII and secrets slip through prompts undetected. When a security team asks "show me every prompt that contained customer data," most teams have nothing to show. SignalVault fixes that. HOW IT WORKSRoute your AI requests through SignalVault using the Python SDK, Node.js SDK, or direct proxy. Your existing OpenAI or Anthropic code works unchanged — just point at a different base URL. Every request is then:• Logged with AES-256-GCM encryption — immutable, tamper-proof audit trail • Scanned for PII — emails, SSNs, phone numbers, credit card numbers, IP addresses • Checked for secrets — API keys, bearer tokens, AWS credentials, connection strings • Evaluated against guardrail rules with configurable actions: block, warn, or redact • Tracked for token usage and costIntegration takes minutes, not days. No code rewrites, no proprietary SDK lock-in. KEY FEATURESEncrypted audit logging: Every prompt and response stored with AES-256-GCM via Cloak. API keys hashed with HMAC-SHA256. Raw keys and unencrypted payloads are never persisted. Export-ready for SOC2 and GDPR reviews.PII detection: Pattern-based detection catches emails, phone numbers, SSNs, credit card numbers, and more before they reach the AI provider. Configure rules to block the request, warn and log, or redact the sensitive content automatically.Secret scanning: Detects API keys, bearer tokens, AWS keys, private keys, and connection strings leaking through prompts. Same block/warn/redact actions.Budget controls: Set monthly cost limits and daily token caps per app. Receive alerts via email or webhook when thresholds are approaching. A budget.exceeded event fires when limits are hit.Guardrail rules: Core rules (PII, secrets, budgets) included on all plans. Growth plan adds custom rules — write your own guardrail policies with pattern matching to enforce what your AI can and can't do.Mirror Mode: Async logging that stays completely off the request path. Zero added latency. Shows what would have been caught without blocking anything. Useful for teams that want visibility before turning on enforcement.Compliance exports: One-click CSV and JSON audit exports included on all plans — not gated behind enterprise pricing. Ready for SOC2, GDPR, HIPAA, and security reviews.Email and webhook alerts: Get notified immediately when policy violations occur, budgets are exceeded, or AI errors spike. POST to Slack, Teams, Discord, or any HTTPS endpoint.Unlimited logging: No log caps on any tier. Every request is recorded, always. Competitors cap logs and stop recording when limits are hit — SignalVault never stops watching. WHAT MAKES SIGNALVAULT DIFFERENTMost AI gateways focus on routing and model switching. SignalVault focuses on trust and compliance — the part that matters when a CISO asks for an audit trail. Three things set it apart:Unlimited logging on every plan. Competitors like Portkey cap at 3M logs on their Pro tier and stop recording. SignalVault logs everything, always.Compliance exports included from day one. Other tools gate CSV/JSON audit exports behind $5K+/month enterprise plans. SignalVault includes them on all tiers.Built for reliability. The entire stack — proxy, detection engine, rule system, and dashboard — is pure Elixir running on the BEAM. No JavaScript frameworks, no Python workers. The BEAM was built for telecom-grade fault tolerance and high concurrency, which is exactly what a proxy needs. TECH STACKRuntime: Elixir / BEAMFramework: Phoenix LiveViewEncryption: AES-256-GCM via Cloak Key hashing: HMAC-SHA256 API format: OpenAI-compatible Database: PostgreSQL PRICINGPer-app subscriptions. Each app gets its own plan.Starter ($49/month): Production environment, 30-day retention, core rules, unlimited logging, compliance exports.Growth ($149/month): All environments (dev, staging, production), 90-day retention, custom rules, email + webhook alerts, Mirror Mode, priority support.Free to try with 100 requests, no credit card required.Built by a solo founder with a background in professional security tooling.

Waxell is the governance plane for agentic systems in production. Agents make real decisions with real consequences — without governance, failures compound silently, costs go unchecked, and compliance gaps widen. Waxell provides two things most agent infrastructure lacks: full observability into what agents are doing — every decision, tool call, and reasoning step — and governance over what they're allowed to do. Enforce policies before, during, and after execution — warn, block, or redact in real time. Deterministic control over probabilistic behavior. MCP-native. Competitors: LangSmith, Helicone, Arize Phoenix, Braintrust.

CheatProof.ai is an AI-powered interview platform designed to detect and prevent AI cheating during technical assessments. Key features include:Real-time AI Detection: Identifies AI use during coding tasks with high accuracy.Secure Online IDE: Provides a collaborative coding environment for candidates to complete tasks.Candidate Activity Tracking: Monitors candidate behavior and timing to identify suspicious patterns.Ideal for:HR & Recruiters: Ensuring fair and authentic candidate evaluations.Tech Leads & CTOs: Analyzing coding behavior and identifying AI-assisted solutions.Startups & Enterprises: Reducing hiring risk and maintaining talent quality.CheatProof.ai offers a multi-signal model tuned with real-world interview telemetry, providing calibrated thresholds to reduce false positives and ensure reliable AI cheating detection.