agent-call — security grade CAUTION, quality 58/100

Security audit verdict: CAUTION · quality 58/100

Flagged: opens a tunnel service. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →

by XiyaoWang0519 · MCP Server · ★ 97

Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h

🔒 Is agent-call safe to install? View the security audit →

About agent-call

Agent Call Agent Call lets an AI assistant make phone calls on your behalf. Give it a task, review the call plan, and explicitly confirm before it dials. You can follow the call, provide answers, stop it, and retrieve the result and transcript. Developed with Codex, Agent Call uses OpenAI GPT-Live SIP () with a delegated task backend for the phone conversation and Twilio to reach the recipient. You run the service yourself, using your own OpenAI and Twilio accounts. Works with ChatGPT Work and Claude web, verified with real question-and-answer calls. It does not work in regular ChatGPT mode. See client setup, test results, and verification limits. What you need | OpenAI API

ai-agentsfastapimcpopen-sourceopenai-realtimepythontelephonytwiliovoice-agent

Quick Facts

Stars97
Forks1
LanguagePython
CategoryMCP Server
LicenseMIT
Quality Score58.4560199024289/100
Open Issues6
Last Updated2026-09-15
Created2026-07-13
Platformsmcp, python
Est. Tokens~13k

Compatible Skills

These tools work well together with agent-call for enhanced workflows:

  • Patter — semantic(0.38)+complementary+shared_fw(openai)+rare_topics+same_lang+shared_platform (70%)

agent-call alternative? Top 6 similar tools

Looking for a agent-call alternative? If you're comparing agent-call with other mcp server tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.

  • OmoiOS by kivo360 · ⭐ 78

    Turn feature specs into merged PRs with a self-supervising swarm of coding agents — parallel execution, isolat

  • omnicoreagent by omnirexflora-labs · ⭐ 250

    The governed runtime for Python agents you can let act: policy on every action, durable runs, sandboxes, budge

  • hermes-talk by TheSmokeDev · ⭐ 57

    Hermes Talk: realtime duplex voice for Hermes Agent. Rides your ChatGPT or X Premium subscription, no API key

  • sandstorm by tomascupr · ⭐ 448

    Run Claude agents in secure cloud sandboxes — via API, CLI, or Slack. One call. Full agent. Zero infrastructur

  • daymon by daymonio · ⭐ 365

    Daymon puts your favorite AI to work 24/7. It schedules, remembers, and orchestrates your own virtual team. Fr

  • orchestkit by yonatangross · ⭐ 286

    The Complete AI Development Toolkit for Claude Code. 106 skills, 36 agents, 171 hooks. Install `ork` for stabl

More MCP Server Tools

Explore other popular mcp server tools:

View all MCP Server tools →

Popular Python Agent Tools

Frequently Asked Questions

What is agent-call?

agent-call is Open-source MCP server for safe, stateful AI-agent phone calls with OpenAI Realtime and Twilio.. It is categorized as a MCP Server with 97 GitHub stars.

What programming language is agent-call written in?

agent-call is primarily written in Python. It covers topics such as ai-agents, fastapi, mcp.

How do I install or use agent-call?

You can find installation instructions and usage details in the agent-call GitHub repository at github.com/XiyaoWang0519/agent-call. The project has 97 stars and 1 forks, indicating an active community.

What license does agent-call use?

agent-call is released under the MIT license, making it free to use and modify according to the license terms.

What are the best alternatives to agent-call?

The top alternatives to agent-call on Agent Skills Hub include OmoiOS, omnicoreagent, hermes-talk. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.

How this security grade is produced

Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.

The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.

Sources & who's responsible:

View on GitHub → Browse MCP Server tools