Agent runtime guard — full-stack runtime interception for AI agents (Agent 运行时全栈拦截). A zero-dependency MCP server that blocks destructive commands, credential theft, SSRF and injection in real time, with Ed25519 evidence receipts. Works with Claude Code,

Secure runtime for AI agents. Run AI agents and their tools in Firecracker microVMs with network policies, PII scanning, prompt injection defense, and audit logging. 7 tools: create_sandbox, run_command, read_file, write_file, list_files, kill_sandbox, list_sandboxes. Security presets (none / standard / strict) let you control guardrails per sandbox: restrict outbound traffic to specific domains, redact PII from command output, block prompt injections, and get full audit trails.

MCP server + instrumentation layer that lets an external agent observe and play Hearts of Iron IV like a player: live memory reads, offline save parsing, and commands injected into the game's own player command queue.

MCP server that injects secrets into shell commands at exec time — plaintext never reaches the model context

MCP server that gates vault credentials behind Chat approval and executes commands with secrets injected locally, so the AI agent never sees the credential.

Use a secret without holding it: an encrypted vault whose values are injected into the commands an agent runs and redacted from everything that comes back.

Give AI agents the use of your secrets without the values: the agent asks for a secret by name, sealkeep injects it into one command and redacts it from the output. Secrets live in the OS keyring, Vault or OpenBao, AWS, Google, Azure, 1Password or Bitwarden.

API keys in the macOS Keychain for Claude Code: a menu bar app with an MCP server that injects keys into commands and redacts them from output

MCP server for remote SSH operations -- persistent sessions, structured command output, SFTP file transfer, port forwarding, and secret-safe environment variable injection with automatic output redaction.

Local-first developer secrets manager — store encrypted env vars, manage API keys across projects, resolve provider identities, and inject secrets into commands. CLI + MCP server. No cloud, no account.

Give AI agents the use of your secrets without the values: named secrets in the OS keyring, with Vault as a replica, injected into a command and redacted from its output. CLI + MCP + guard hook.

MCP server for live Qt UI automation. Inject into running QWidget or QML/QQuickWindow apps, query the live tree, act/get/set/call, screenshot/annotate, and resolve QML source. Qt 5/6; Windows/Linux.

Anti-hallucination MCP server for AI agents. 4-gate dual-stage pipeline. Go binary. Blocks fabricated claims, identity drift, credential leaks, dangerous commands.

MCP-native SSH bridge for AI agents with host-scoped access, rule-based command validation, ephemeral SSH sessions, and audit logging

What a checkout you did not write tells your agent to do — instruction files, hook commands, declared MCP servers, and repo-shipped skills. Read-only MCP server.

MCP guard & vibe coding security tool for Cursor and Claude Code. Fail-closed human-in-the-loop approval for shell commands and MCP tools—local daemon, YAML policy, no cloud.

MCP server that shows what a checkout you did not write tells your agent to do — instruction files, hook commands wired to agent events, MCP servers it declares, and the skills and subagents it ships.

Deliberately vulnerable MCP server (aka MCP Goat) for security training — 26 challenges across 4 difficulty levels (incl. a secure reference), a victim-agent harness, and one-command Docker deploy. Practice penetration testing against the Model Context Protocol.