Marketplace / Blueprint 08
LIVEAgents / control

Secure Coding Agent

Resume and operate coding agents remotely while humans retain approval over consequential actions.

Review my build

Interactive platform map

Architecture in context

The focused blueprint, its required foundation, and declared recommendations.

Core Selected Automatic Required path
Blueprint 08 / Agents

Secure Coding Agent

Resume and operate coding agents remotely while humans retain approval over consequential actions.

2 vCPU4 GB RAM6 services
01 / Problem

What this replaces

Autonomous coding agents are valuable, but unrestricted file writes and shell commands create unacceptable operational risk.

02 / Outcome

What your team gains

A remote coding experience exposed through chat/Telegram with explicit approve-or-deny checkpoints for writes and shell operations.

03 / Capability

What is inside the blueprint

Resume Claude Code sessions from chat
Telegram coder bot
OpenWebUI CC Sessions pipe
Human approval gates for writes
Human approval gates for shell commands
Remote session control
Integration with persistent session memory and graph history
Multi-channel delivery of agent requests
04 / Architecture

How it fits the platform

Open WebUI or Telegram -> coder gateway -> Claude Code session -> approval gate -> filesystem/shell action -> session history/telemetry.

Included services

telegram-coder, owui-coder, telegram-gateway, Open WebUI, session-recall-mcp, Neo4j stack

Platform requirements
05 / Delivery

From prerequisites to operation

Prerequisites
  1. Claude Code/runtime environment
  2. Telegram bot credentials
  3. Session persistence
  4. Defined approval policy
Deployment
  1. Deploy coder gateway modes
  2. Connect OpenWebUI pipe and Telegram bot
  3. Map sessions
  4. Define gated operations
  5. Integrate session recall/graph optionally
Configuration
  1. Approval scope
  2. Allowed commands/actions
  3. Session mapping
  4. Bot access lists
  5. Timeout and retry behavior
Operations
  1. Session cleanup
  2. Secret management
  3. Gateway health
  4. Auditability of approved actions
  5. Dependency compatibility
06 / Combinations

What this unlocks with other layers

MCP Agent Hub + Secure Coding Agent

Controlled Agency

Agents can act across systems while risky writes and commands remain human-gated.

07 / Technology

Technology behind this capability

Open WebUIRUNNING - v0.11.0 in censusNeo4jRUNNING - 5-community in censusModel Context Protocol (MCP)MULTIPLE RUNNING SERVERSsession-recall-mcpAICORTEX nativetelegram-gatewayAICORTEX nativetelegram-coderAICORTEX nativeowui-coderAICORTEX native