Skip to content
ATAI Today Brief
HomeNewsConceptsGuidesToolbox
AboutSubscribeUA
Subscribe

AI Today Brief

The daily AI-engineering brief. Built in public. EN · UA.

XTelegramLinkedInYouTubeRSS

Follow AI Today Brief on LinkedIn for daily AI-engineering updates and the weekly “5 shifts that changed how developers work” PDF.

Explore

NewsDigestsConceptsGuides

Company

SubscribeAdvertiseAbout

Legal

Editorial policyAI disclosurePrivacyTerms

© 2026 AI Today Brief. All rights reserved.

  1. Home/
  2. News/
  3. Agents & MCP/
  4. OneCLI: Open-Source Gateway Keeps Real Credentials Out of AI Agents
Agents & MCP

OneCLI: Open-Source Gateway Keeps Real Credentials Out of AI Agents

July 23, 2026· 5 min read
OKCurated by Oleksandr Kuzmenko, AI Product Engineer·Updated July 23, 2026·Sources cited on every story
AI-assisted · editor-reviewed·How we use AI
OneCLI: Open-Source Gateway Keeps Real Credentials Out of AI Agents

OneCLI is an open-source Rust proxy that sits between AI agents and target services to inject real API keys dynamically. Agents interact using placeholder tokens, while OneCLI handles AES-256-GCM secret decryption and pattern-based header replacement. This prevents autonomous agent tools from ever exposing or exfiltrating raw credentials.

Impact: High

Why it matters

You can securely give autonomous agents access to dozens of APIs without risking production key leakage.

TL;DR

  • 01Store API keys in an encrypted gateway, giving agents placeholder tokens like FAKE_KEY.
  • 02Intercept outbound HTTP/HTTPS requests on port 10255 to swap keys transparently by host and path.
  • 03Integrate Bitwarden or local secret stores to prevent raw credential disk persistence.

Key facts

Gateway Port10255
Dashboard Port10254
Gateway Port
10255
Dashboard Port
10254
Encryption Standard
AES-256-GCM
Auth Modes
Single-user local, Google OAuth

Zero-Trust Proxy Architecture for Agent Workflows

Giving agentic workflows direct access to production secrets risks credential exfiltration during sandbox escapes or unexpected model behavior. OneCLI shifts authentication responsibility to a fast Rust-based HTTP proxy layer, ensuring agent tools never process raw API keys directly.

Transparent Credential Swap via Host and Path Matching

Instead of hardcoding environment variables, agents configure standard HTTP request calls using dummy access tokens (e.g., FAKE_KEY). When the agent sends a request through localhost:10255, OneCLI intercepts the headers, resolves the host and path pattern, decrypts the true key at runtime, and injects it into outbound headers or URL parameters.

Local-First Deployment with Docker Compose

OneCLI can be deployed locally using a single shell installer or Docker Compose. It includes built-in password manager integration (such as Bitwarden) for fetching keys on demand without storing secrets on disk, alongside optional Google OAuth for multi-developer team environments.

Try it in 2 minutes

curl -fsSL https://onecli.sh/install | sh
# Dashboard: http://localhost:10254
# Rust Gateway Proxy: http://localhost:10255

bash

✓ When to use

  • Deploying autonomous agents (Claude Code, Cursor agent mode) with access to paid third-party APIs.
  • Managing multi-agent environments where granular, per-agent credential scopes are required.

✕ When NOT to use

  • Simple local scripts where tool calls do not touch external HTTP APIs.
  • Environments where client network traffic cannot be routed through a local HTTP proxy.

What to do today

  • →Run `curl -fsSL https://onecli.sh/install | sh` to boot the local dashboard and gateway.
  • →Configure your Claude Code, Cursor, or custom AI agent HTTP proxy settings to point to `http://localhost:10255`.

What the community says

  • “Oo this is interesting, blocking it on the network level is nice”

    — redbridgerock on Hacker News

  • “It works great for static credentials... Gateway does the OAuth dance to get the access token, this is quite practical”

    — sandeepkd on Hacker News

#OneCLI#Rust#Next.js#Docker#Bitwarden

Sources

  • OneCLI GitHub Repository
  • Hacker News OneCLI Discussion
ShareShare on XShare on LinkedIn
← Previous storyClaude Pro Users Warned Against Automatic Billing Trigger on Free Fable CreditsNext story →Runway Launches Media Router API for Dynamic Generative Model Selection

Related stories

  • Agents & MCPAnalyzing AI Agent Sandbox Escape Mechanisms via Package Proxy Vulnerabilities
  • Agents & MCPGoogle Study Shows Active Agent Questioning Outperforms Unprompted LLMs in Diagnostic Accuracy
  • Agents & MCPOpenAI Launches Presence for Reliable AI Agent Deployment and Orchestration
  • Agents & MCPAwesome LLM Apps: Over 100 open-source agent templates and skills to run instantly

Email digest

Get the morning AI brief

One email a day — the stories that matter for engineers, founders and tech leads. Human-edited, with links to primary sources.

  • ✓120+ sources scanned daily
  • ✓Edited by a human
  • ✓1 email per day
  • ✓EN + UA

By subscribing you agree to the privacy policy.