claude-code

Delegate coding tasks to the Claude Code CLI via print mode or tmux interactive sessions.

14|5|Updated Apr 9, 2026
One-click install
npx skills add https://github.com/MLT-OSS/hermes-agent-go --skill claude-code-mlt-oss
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: claude-code
Source: https://github.com/MLT-OSS/hermes-agent-go/tree/main/skills/autonomous-ai-agents/claude-code
Command: npx skills add https://github.com/MLT-OSS/hermes-agent-go --skill claude-code-mlt-oss

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @anthropic-ai/claude-code.

What problem does it solve? Orchestrating an autonomous coding agent from within another agent requires knowing the right CLI flags, handling interactive permission dialogs, and choosing between one-shot and multi-turn execution modes. This Skill provides the complete operational playbook for driving Claude Code from Hermes without trial and error. ## Core Features & Use Cases - Print Mode Automation: Run one-shot coding tasks with claude -p, structured JSON output, JSON schema extraction, session resumption, and budget/turn limits for CI pipelines. - Interactive PTY Orchestration: Drive multi-turn Claude Code sessions through tmux, including handling workspace trust and permission-bypass dialogs with send-keys sequences. - PR Review & Parallel Workflows: Review diffs and PRs by number, run isolated git worktrees, and execute multiple independent Claude tasks in parallel tmux sessions. - Use Case: You need to refactor an authentication module and add tests. Launch Claude Code in print mode with --allowedTools 'Read,Edit' --max-turns 10, capture the JSON result with cost and session ID, then resume the session to iterate. ## Quick Start Ask the agent to run claude -p with your coding task, a working directory, and a max-turns limit to get a one-shot autonomous code change.

Frequently Asked Questions about claude-code

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run Claude Code non-interactively for automation?

Use print mode with claude -p followed by your task prompt. It runs one-shot, skips all interactive dialogs, and exits when done. Add --max-turns to cap agentic loops and --output-format json for structured results.

How do I handle Claude Code permission dialogs in tmux?

The workspace trust dialog defaults to 'Yes', so send Enter via tmux send-keys. The bypass-permissions warning defaults to 'No', so send Down then Enter to select 'Yes, I accept' before the session proceeds.

Can Claude Code output structured JSON for scripting?

Yes, use --output-format json to get a result object with session_id, num_turns, total_cost_usd, and the result text. You can also pass --json-schema to force output matching a specific schema, or stream-json for real-time events.

How do I resume a previous Claude Code session?

Use claude -c to continue the most recent conversation in the current directory, or claude -r with a session ID from a prior JSON result. Add --fork-session to branch the history into a new session ID.

What are the limitations of Claude Code print mode?

Print mode is single-shot with no multi-turn conversation, so iterative tasks need session resumption via --resume or --continue. Flags like --max-turns and --max-budget-usd only apply in print mode, and --fallback-model is print-only.