How do I install the Call Claude skill?

Run npx skillmds@latest add stablyai/call-claude in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Call Claude safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Call Claude?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Call Claude free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under MIT.

Who published Call Claude?

stablyai (@stablyai) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Call Claude by stablyai — Agent Skill

Name: Call Claude
Author: stablyai

Call Claude

Delegate a prompt to Claude via `claude -p`. Use when the user wants to offload a task to Claude from another agent (Codex, Cursor, another Claude session), get a second opinion, or run a parallel implementation. Default model is opus.

stablyai ec5512f 2 files · 2.7 KB Updated Sep 17, 2026

File contents

call-claude

Delegate an arbitrary prompt to Claude and return the output.

Arguments

Required: a prompt.

Optional flags (parsed from the user's message):

Flag	Values	Default
`--model`	`opus`, `sonnet`, `haiku`, or a full model ID	`opus`
`--effort`	`low`, `medium`, `high`, `xhigh`, `max`	(omitted)

Process

Parse the user's message — extract the prompt and any --model / --effort overrides.

Invoke the wrapper script that ships alongside this SKILL.md:

"$(dirname "$0")/call-claude.sh" [--model <alias>] [--effort <level>] "<prompt>"

For multi-line prompts, pipe via stdin:

echo "<prompt>" | "$(dirname "$0")/call-claude.sh" --model opus

Return Claude's full output to the user. Don't summarize unless asked.

claude -p is non-interactive and prints the final answer to stdout, so the wrapper output can be captured directly.

Prerequisite

Claude Code must be installed — verify with claude --version.

Examples

# Default (opus)
./call-claude.sh "explain what this repo does"

# Override model
./call-claude.sh --model sonnet "summarize the README"

# Override model + effort
./call-claude.sh --model opus --effort high "design a caching layer"

Frequently asked questions