SkillAgentSearch skills...

agentstack

One agent setup across every coding CLI — import, render, switch toolsets, diagnose drift, and restore safely. Local-first, vendor-neutral, Rust.

Install / Use

claude mcp add Tarekkharsa -- npx -y github:Tarekkharsa/agentstack

If the server publishes to npm under a different name, use that package instead — check the repo README.

About this skill
🔌

MCP Server

Model Context Protocol server

Quality Score

70/100

Supported Platforms

Claude Code
Claude Desktop
Cursor
Gemini CLI
OpenAI Codex

Our assessment of agentstack

agentstack scores 70/100 on our quality scale, 3799th of 4,592 Development & Engineering skills we index.

Its MCP Server is 4.0 KB long, split into 4 sections with 1 code example: a solid amount of guidance for an agent.

It has 3 GitHub stars, so there is little community track record yet; judge it on its content.

Substance
26/30
Structure
12/20
Description
15/15
Adoption
3/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated about 2 months ago, so agentstack is actively maintained.
  • Our last check on 2026-09-23 found the source still online.
  • It is released under the Apache-2.0 license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 92/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands (1 minor note below). An AI review of the same text found nothing harmful.

  • noteInstalls by piping a downloaded script into a shellline 17
    curl -fsSL https://raw.githubusercontent.com/Tarekkharsa/agentstack/main/install.sh | sh

AI review by kimi-k2.7-code on 2026-10-05. Automated pattern scan on 2026-10-06. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

agentstack compared with similar skills

All 4 of these similar skills score higher than agentstack; compare them before choosing.

SkillScoreStarsUpdatedFormat
agentstack (this skill)by Tarekkharsa70358d agoMCP Server
Agent-Reachby Panniantong10095.9k3d agoCLAUDE.md
headroomby headroomlabs-ai10075.1ktodayCLAUDE.md
CowAgentby zhayujie10047.3ktodayCLAUDE.md
ai-job-searchby MadsLorentzen10046.1k3d agoCLAUDE.md

Frequently asked questions

How do I install agentstack?
Run claude mcp add Tarekkharsa -- npx -y github:Tarekkharsa/agentstack. The install tabs above show the steps for each supported agent.
Which AI agents does agentstack work with?
It is written for Claude Code, Claude Desktop, Cursor, Gemini CLI and OpenAI Codex, as a MCP Server file. Other agents that read the same format can often use it too.
Is agentstack safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands (1 minor note below). An AI review of the same text found nothing harmful. It is Apache-2.0-licensed and scores 92/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is agentstack still maintained?
The repository was last updated about 2 months ago, so agentstack is actively maintained.
<img alt="agentstack" src="docs/logo.svg" width="380">

One agent setup. Every coding CLI. Today you configure the same MCP server once per tool, in a different format each time, with your tokens sitting in plain JSON. With AgentStack the whole setup — config, skills, servers, instructions — lives in one repo you own, any machine gets it with one command, and every CLI speaks it: Claude Code, Codex, Cursor, Gemini CLI, OpenCode and eight more — served live where the tool speaks MCP, written as native files where it does not.

CI Conformance Release License

Try it in 60 seconds

curl -fsSL https://raw.githubusercontent.com/Tarekkharsa/agentstack/main/install.sh | sh
export PATH="$HOME/.local/bin:$PATH" # only if the installer printed "Add to PATH"
agentstack init --connect            # import what your CLIs already have, and connect them
# now restart your coding CLI — it reads its config at startup
agentstack status                    # is it ready — and if not, the one thing that fixes it
agentstack more why <server>         # which of your CLIs is served this, and from where

The installer verifies the download against the checksums published with the release. Plain init is a guided wizard and needs a terminal; init --yes is the scripted form. more is the extended toolbox, and every more command also runs at its bare name.

Two more worth knowing early: agentstack doctor tells you what is wrong and what fixes it, and agentstack more restore takes any write back.

Two CLIs with different half-setups: agentstack imports both into one setup file, connects them so both are served the servers live while the project stays clean, passes doctor with 0 errors, writes each native format on request, and restores the machine byte-for-byte

Your second machine

Publish your setup into a library repo you own, commit it, and the next machine is one command: agentstack up --library <git-url> --write. Two things stay per machine by design: secret values, and the trust review — you read what the project declares, and say yes once. → Set up another machine

Where to go next

Supported platforms: macOS and Linux. A Windows binary is published but is not exercised by CI — treat it as untested rather than supported.

MIT OR Apache-2.0.

Related Skills

View on GitHub
GitHub Stars3
CategoryDevelopment
Updated1mo ago
Forks1

Languages

Rust

Trust signals

92/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 low