SentryFrogg MCP Server
Provides a governed command plane for MCP agents to manage PostgreSQL databases, SSH sessions, and HTTP APIs with encrypted credential storage and prescriptive workflows.
README
SentryFrogg MCP Server v4.2.0
Document Profile
| Field | Value |
|---|---|
| Product | SentryFrogg MCP Server |
| Version | 4.2.0 |
| Runtime | Node.js ≥ 16 |
| Interfaces | Model Context Protocol (PostgreSQL · SSH · HTTP) |
| License | MIT |
Executive Synopsis
SentryFrogg MCP Server supplies a governed command plane for MCP-compatible agents that operate data platforms, remote shells and web services. The implementation prioritises deterministic behaviour, encrypted credential storage and prescriptive workflows so that autonomous agents can execute infrastructure tasks with enterprise auditability.
Functional Surface
| Capability | Detail |
|---|---|
| PostgreSQL control | Parameterised SQL, catalog discovery, CRUD helpers, mutual TLS, connection profiling. |
| SSH execution | Serial command dispatch per profile, password/key authentication, reachability diagnostics. |
| HTTP access | REST verbs with JSON payloads, health checks, programmable headers and bearer tokens. |
| Secret vault | AES-256 encrypted profiles persisted in profiles.json with managed key material. |
| Telemetry | Tool-level statistics via getStats() for integration with monitoring stacks. |
System Components
| Component | Scope |
|---|---|
sentryfrogg_server.cjs |
MCP entry point, tool catalogue, lifecycle supervision. |
src/bootstrap/ServiceBootstrap.cjs |
Service registration, dependency wiring, resource cleanup. |
src/managers/PostgreSQLManager.cjs |
SQL execution, profile validation, TLS configuration, pool management. |
src/managers/SSHManager.cjs |
SSH sessions, sequential execution enforcement, profile hygiene. |
src/managers/APIManager.cjs |
HTTP invocation, header synthesis, response shaping. |
src/services/ProfileService.cjs |
Encrypted persistence and retrieval of profile objects. |
src/services/Security.cjs |
Key lifecycle, cryptographic primitives, payload size guards. |
src/services/Validation.cjs |
Canonical validation for incoming payloads. |
MCP Tools
mcp_psql_manager
| Attribute | Specification |
|---|---|
| Required flow | setup_profile → downstream action using same profile_name. |
| Actions | setup_profile, list_profiles, quick_query, show_tables, describe_table, sample_data, insert_data, update_data, delete_data, database_info. |
| Credentials | Either discrete fields (host, port, username, password, database) or connection_url. |
| TLS options | ssl_mode, ssl_ca, ssl_cert, ssl_key, ssl_passphrase, ssl_servername, ssl_reject_unauthorized; omitted values reuse stored secrets. |
| Query rules | quick_query injects LIMIT 100 when absent; bind variables supplied via params array ($1, $2, ...). |
| Response format | JSON with success, rows, rowCount, fields, command; failures raise MCP internal errors. |
| Rejection triggers | Missing profile, malformed SQL, payload limits exceeded, invalid TLS configuration. |
mcp_ssh_manager
| Attribute | Specification |
|---|---|
| Required flow | setup_profile (password or PEM private_key, optional passphrase) → operational action. |
| Actions | setup_profile, list_profiles, execute, system_info, check_host. |
| Execution model | Commands trimmed and length-limited; pipes/redirects permitted; per-profile execution is strictly sequential. |
| Outputs | JSON containing success, stdout, stderr, exitCode, durationMs; errors propagate as MCP internal errors. |
| Security posture | Secrets encrypted at rest; no templating—agents must supply fully qualified commands. |
mcp_api_client
| Attribute | Specification |
|---|---|
| Actions | get, post, put, delete, patch, check_api. |
| Inputs | url (required), data (JSON body for mutating verbs), headers (string map), auth_token (prefixed into Authorization unless already set). |
| Behaviour | Local and private addresses allowed; HTTP status/body returned in structured JSON; transport or parsing failures emit MCP internal errors. |
Profile Lifecycle
- Invoke
setup_profileto persist credentials and TLS artefacts; secrets encrypt with AES-256 using.mcp_profiles.key(0600permissions). - Reference the same
profile_namefor subsequent operations; omitted sensitive fields inherit stored encrypted values. - Rotate credentials by reissuing
setup_profile; the latest payload supersedes previous entries. - Audit existing profiles via
list_profiles; responses never disclose secrets. - Retire unused profiles by editing
profiles.jsonunder change control.
TLS Configuration Guidance
- Prefer embedding
sslmodedirectives inconnection_url; explicit payload fields override URL parameters. - Keep
ssl_reject_unauthorizedattrueunless communicating with trusted self-signed endpoints. - Provide
ssl_servernamewhenever certificate CN/SAN mismatches the host. - Supply PEM blocks as single-line strings using
\nescape sequences; leading/trailing spaces are disallowed. ssl_passphrasemust be non-empty if provided; omit otherwise.
Installation and Operations
| Task | Command |
|---|---|
| Clone and install | git clone https://github.com/yourusername/sentryfrogg-mcp.git && cd sentryfrogg-mcp && npm install |
| Syntax check | npm run check |
| Launch (stdio) | node sentryfrogg_server.cjs |
| Update dependencies | npm install --package-lock-only && npm audit fix --only=prod (subject to governance) |
| Reset profile store | Remove profiles.json after confirming backups |
Security & Compliance
- Encryption key lifecycle:
.mcp_profiles.keygenerated on first run; override viaENCRYPTION_KEYfor coordinated environments. - Secret exposure: MCP responses never include decrypted values; rotation requires explicit
setup_profile. - Input governance: SQL statements, SSH commands and HTTP payloads are length-limited; oversized inputs are rejected pre-execution.
- Audit trail: stderr logging captures timestamped events per tool to support collection by SIEM platforms.
- Dependency governance: locked versions of
pg,ssh2,node-fetch,@modelcontextprotocol/sdk; monitor advisories for patch cadence.
Troubleshooting Matrix
| Symptom | Diagnostic Actions | Remediation |
|---|---|---|
| PostgreSQL TLS failure | Inspect ssl_mode, ssl_servername, certificate chain, Postgres logs. |
Update TLS materials; rerun setup_profile. |
| SSH command hang | Validate command length, ensure non-interactive execution, check remote prompts. | Adjust command or script; rerun setup_profile if credentials changed. |
| HTTP error response | Review returned status/body, verify headers and auth_token. |
Correct payload; retry request. |
| Missing profile | Execute list_profiles to confirm presence; ensure consistent profile_name. |
Recreate via setup_profile. |
Change History Reference
Consult CHANGELOG.md for a dated record of functional and operational updates, including TLS support and renaming.
Contribution & Support
- Submit changes through pull requests accompanied by verification evidence (
npm run check). - Never commit
.mcp_profiles.keyor environment-specific secrets. - Use maintainer contact information in
package.jsonfor escalation or integration assistance.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.