temple-bridge

Temple Bridge is a local MCP (Model Context Protocol) server that exposes tool-based access (8 tools, 3 resources) for an LLM-based workflow. It appears designed to connect an "Action Layer" repository (back-to-the-basics) with a "Governance/Memory" repository (threshold-protocols), using middleware to maintain state across MCP tool calls and a human-in-the-loop approval gate for command execution.

Evaluated Mar 30, 2026 (21d ago)
Repo ↗ DevTools mcp fastmcp local-ai tool-calling python agentic-workflows governance middleware lm-studio mlx
⚙ Agent Friendliness
50
/ 100
Can an agent use this?
🔒 Security
47
/ 100
Is it safe for agents?
⚡ Reliability
32
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
70
Documentation
65
Error Messages
0
Auth Simplicity
55
Rate Limits
0

🔒 Security

TLS Enforcement
60
Auth Strength
30
Scope Granularity
35
Dep. Hygiene
65
Secret Handling
55

Security is primarily based on local sandboxing (BTB path restriction), an allowlist of safe commands, and an interactive approval gate for command execution in LM Studio. No authentication for the MCP server is described. TLS/network transport security cannot be confirmed from the provided materials. Dependencies are limited per manifest (fastmcp, python-dotenv), but CVE status and secure configuration practices are not assessable from the given data.

⚡ Reliability

Uptime/SLA
0
Version Stability
45
Breaking Changes
40
Error Recovery
45
AF Security Reliability

Best When

You want a local, MCP-compatible agent that can perform controlled repository operations with an explicit human approval step before executing commands.

Avoid When

You need a standardized, externally hosted API with strong auth, documented rate limits, or you cannot reliably enforce/monitor local sandboxing and approval behavior.

Use Cases

  • Local LLM tool-calling with a governance/memory layer
  • Running and reviewing tests or other bounded developer tasks in a sandboxed repository
  • Reading and exploring repository files via an agent (list/read with path restrictions)
  • Human-approved code execution workflows ("threshold witness")
  • Stateful multi-step agent sessions with phase tracking (Spiral protocol)

Not For

  • Unattended automation of arbitrary shell commands
  • Use cases requiring strong enterprise authentication/authorization or audit guarantees beyond local machine controls
  • Production deployments where network security, remote access, and SLAs are critical

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: Human-in-the-loop approval in LM Studio for command execution (described as 'Threshold Witness')
OAuth: No Scopes: No

No user authentication/authorization mechanism is described for the MCP server itself; security appears to rely on local process control, allowlisted commands, path sandboxing, and interactive approval.

Pricing

Free tier: No
Requires CC: No

Project is MIT-licensed; README describes local operation with an on-device model (e.g., Hermes via MLX/LM Studio), so costs are primarily local compute and any hardware/software tooling.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Command execution is gated by interactive approval; agents running without a human present may stall.
  • Tool timeouts are described (default 60s; configurable), so long-running commands may fail without retries.
  • The server relies on correct local configuration paths (TEMPLE_BASICS_PATH / TEMPLE_THRESHOLD_PATH) and LM Studio MCP connectivity.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for temple-bridge.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered