wundervault
OfficialRelated Servers
Alternatives to wundervault
- AlicenseAqualityDmaintenanceEnables interaction with HashiCorp Vault to read, write, list, and delete secrets through a containerized MCP server with secure token-based authentication.4MIT
- AlicenseNot gradedqualityDmaintenanceMCP server that lets AI agents call APIs without ever seeing the credentials, using a local encrypted vault and per-secret allowlist policies for HTTP requests and subprocess environment variables.1AGPL 3.0

AgentValetofficial
AlicenseAqualityCmaintenanceIdentity and credential governance for AI agents. Every agent gets its own cryptographic identity, scoped short-lived credentials per platform, human approval on sensitive actions, and an immutable audit log.72MIT
Related Servers
- AlicenseNot gradedqualityDmaintenanceMCP server enabling AI agents to use secrets (API keys, tokens) via encrypted vault, executing HTTP/shell/SSH actions server-side while never exposing secret values to the AI.MIT

@lobsterkit/vault-mcpofficial
AlicenseAqualityDmaintenanceMCP server for LobsterVault, enabling AI agents to securely store, retrieve, manage, and share encrypted secrets with KMS envelope encryption, including environment variable injection.1114 npmMIT- AlicenseAqualityBmaintenanceMCP server for Vaultwarden/Bitwarden vault management. Enables AI agents to securely create, search, read, and update vault items via the official Bitwarden CLI, with safe-by-default redaction and support for both stdio and SSE transports.5329MIT
- AlicenseNot gradedqualityDmaintenanceSecrets management MCP server that injects credentials into API requests for AI agents, enforcing policies and logging all activity without exposing raw keys.109 npm30MIT
- AlicenseNot gradedqualityBmaintenanceMCP server for DemiPass secrets management, enabling AI agents to securely store, rotate, and use credentials without exposing them in context windows.73 npmMIT
- AlicenseAqualityDmaintenanceA local-first secrets vault for MCP clients that lets AI assistants use secrets by injecting them into child processes, without the plaintext ever entering the model's context.5Apache 2.0
TDQS
Scored across 11 tools
The eleven tools each target a distinct action-resource pair: call a web service, pay a 402, save an existing secret, sign, push secrets to a Worker, mint a new key, get a 2FA code, list entries, git over SSH, exec over SSH, and finish. The main friction is vault_call vs vault_new_key and vault_save_secret vs vault_new_key, since all three end with a secret in the vault, but the descriptions draw explicit boundaries (vault_call refuses the key-creation addresses; save takes a value you already have, new_key mints one at a service). An agent should be able to select correctly without guessing.
Every name shares a single vault_ prefix in snake_case, so the family is instantly recognizable. Most are verb-led (call, pay, sign, save, send, done), with a few noun-form exceptions (vault_git, vault_ssh_exec, vault_entries_list), which is a minor deviation rather than a mixed convention.
Eleven tools for a secrets broker covering HTTP calls, payments, signing, SSH/git, 2FA, key provisioning and lifecycle signalling is well within a reasonable scope. Each tool corresponds to a genuine capability and none looks like filler or a near-duplicate.
The agent-facing surface covers discovery (entries list), all the main secret-using paths (HTTP, pay, sign, 2FA, SSH, git, Worker secrets), provisioning (save, new key) and clean termination (done). Gaps are mostly owner-side by design: no tool to delete/revoke an entry, no way to inspect the audit log, and no direct rotation of an existing secret beyond the new_key replace flow, which agents must work around via vault_call DELETE.