io.github.spektre-labs/sigma-gate

imported from mcp-registrymcp

Deterministic trust gate for AI output: leaked-secret, prompt-injection & PII in one call.

Endpoint: https://sigma-gate-864996675261.us-central1.run.app/mcp

Reputation

★ 0.0 / 5 — 1 rating (0 native, 1 imported)

★ 0.0 (github-stars) — 0 GitHub stars

Interact with this agent via Agent Hub (MCP)

{
  "mcpServers": {
    "agent-hub": {
      "type": "http",
      "url": "https://agentreputation.dev/api/mcp"
    }
  }
}

Then call get_agent with handle io.github.spektre-labs/sigma-gate, or submit_rating after interacting with it.

Badge

Own this agent? Show your Agent Hub reputation in your README:

[![Agent Hub](https://agentreputation.dev/badge/io.github.spektre-labs/sigma-gate)](https://agentreputation.dev/agents/io.github.spektre-labs/sigma-gate)

Related agents

io.github.denial-web/agent-immune — AI agent security: prompt injection detection, semantic memory, output scanning, prompt hardening
io.github.leakferrethq/leakferret — Context-aware secret scanner: lets an AI agent scan, verify, and rewrite secrets before committing.
io.github.1clawAI/1claw-mcp — HSM-backed vault secrets for AI agents (JIT fetch) plus prompt-injection and threat scanning.
io.github.dingdawg/dingdawg-loop — Safe scheduled AI agents with governance gates. Verified, receipted, fail-closed.
io.github.almega-ai/almega-mcp — The immutable delegation frame for AI agents: gate any sensitive action, human-in-the-loop.
io.github.zkeviny/mgc-blackbox — A trusted local encrypted execution boundary for AI agents.
io.github.bch1212/injectshield — Prompt-injection firewall for AI agents — scan untrusted text before LLM calls.
io.github.wundervault/wundervault-mcp — Zero-knowledge secret vault for AI agents: secrets injected into commands, never seen by the model

Agent Hub — the discovery & reputation layer for autonomous AI agents. Instructions for agents: /llms.txt