MCP server for action governance on AI agents — 14 tools that govern tool calls, SQL queries, file writes, shell commands, and agent-to-agent
If you already use Shrike MCP, the shrike mcp mcp server is the piece that lets your assistant work with it directly. MCP server for action governance on AI agents — 14 tools that govern tool calls, SQL queries, file writes, shell commands, and agent-to-agent messages server-side, before execution. Allow / approve / block. Session correlation. Works with.
Shrike MCP is the Model Context Protocol server for Shrike. From employees using ChatGPT to autonomous agents executing code — Shrike evaluates every AI interaction in real-time with tools to scan prompts, responses, SQL queries, file writes, CLI commands, web searches, and agent-to-agent messages. Detects prompt injection, jailbreaks, data leakage, PII exposure, and multi-turn manipulation before they cause harm.
The toolset is worth reading before you wire it up, because it tells you what the integration is really for:
scan_prompt — User/system prompts before LLM processingscan_response — LLM outputs before returning to userscan_sql_query — SQL queries before database executionscan_file_write — File paths and content before writescan_command — CLI commands before shell executionscan_web_search — Search queries before executionscan_a2a_message — Agent-to-agent messages before processingscan_agent_card — A2A AgentCard metadata before trustingscan_mcp_schema — MCP tool definitions before trusting themcheck_approval — Human-in-the-loop approval statusreport_bypass — User-reported missed detectionsreset_session — Clear session correlation stateConfiguration is passed through the environment: SHRIKE_API_KEY, SHRIKE_BACKEND_URL. Treat anything key-shaped as a real credential — scope it to the minimum the server needs, and rotate it if it ever lands in a shared config.
The server ships on npm as shrike-mcp, so your MCP client can launch it on demand — there is no separate build step. Add the server block to your client's configuration, restart it, and the tools register themselves.
This sits in the database access group, where several servers overlap in what they claim to do but differ sharply once you actually set them up. Shrike MCP's toolset — scan_prompt, scan_response, scan_sql_query and 11 more — is a fair guide to whether it matches your workflow. It is maintained by shrike-security; worth a glance at recent repository activity before you build anything load-bearing on it.
This entry was verified against Shrike MCP's own documentation before publication; SyncDev keeps the directory reviewed rather than auto-generated.
| Tool | What it does |
|---|---|
| scan_prompt | User/system prompts before LLM processing |
| scan_response | LLM outputs before returning to user |
| scan_sql_query | SQL queries before database execution |
| scan_file_write | File paths and content before write |
| scan_command | CLI commands before shell execution |
| scan_web_search | Search queries before execution |
| scan_a2a_message | Agent-to-agent messages before processing |
| scan_agent_card | A2A AgentCard metadata before trusting |
| scan_mcp_schema | MCP tool definitions before trusting them |
| check_approval | Human-in-the-loop approval status |
| report_bypass | User-reported missed detections |
| reset_session | Clear session correlation state |
| session_status | Read-only lookup of L9 session state |
| scan_declare_scope | Declared operating scope for task-scoped agents |
{
"mcpServers": {
"shrike-security": {
"command": "npx",
"args": ["-y", "shrike-mcp"]
}
}
}Configuration as documented by the project. Restart the client after saving.
| Variable | Description | Required |
|---|---|---|
| SHRIKE_API_KEY | Credential the server authenticates with. | Yes |
| SHRIKE_BACKEND_URL | Endpoint or connection string the server talks to. | Yes |
Read-only SQL access to Postgres — let your assistant inspect schemas and answer questions from real data.
Manage your whole Supabase project in conversation — database, auth, storage, Edge Functions and branches.
Query, modify and analyse local SQLite databases in conversation — the fastest way to chat with a data file.
Metabase ships its own MCP endpoint — search your BI content, build and run queries, and save questions and dashboards without leaving the chat.
Official MongoDB server covering data, schemas and Atlas management — from find queries to spinning up clusters.
Serverless Postgres with database branching — point your assistant at Neon and let it work on disposable copies.