Claude Code Plugins

Community-maintained marketplace

Feedback

memex-cli

@chaorenex1/coding-workflow
0
0

Execute AI-powered command-line tasks with memory, replay, and resume capabilities using memex-cli. Use when (1) Running AI backend tasks (codex, claude, gemini), (2) Replaying recorded events from JSONL files, (3) Resuming interrupted runs by run_id, (4) Executing prompts with streaming output in jsonl or text format.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name memex-cli
description Execute AI-powered command-line tasks with memory, and resume capabilities using memex-cli. Use when (1) Running AI backend tasks (codex, claude, gemini), (2) Resuming interrupted runs by run_id, (3) Executing prompts with streaming output in jsonl or text format.

Memex CLI

A CLI wrapper for AI backends with built-in memory, and resume support.

Core Commands

Run a Task

memex-cli run \
  --backend <backend> \
  --prompt "<prompt>" \
  --stream-format <format>

Parameters:

  • --backend: AI backend (codex, claude, gemini)
  • --prompt: Task prompt
  • --stream-format: Output format (jsonl or text)
  • --model: Model name (optional, for codex backend)
  • --model-provider: Model provider (optional, for codex backend)

Resume a Run

memex-cli resume \
  --run-id <RUN_ID> \
  --backend <backend> \
  --prompt "<prompt>" \
  --stream-format <format>

Continues a previous run using its run_id.

Backend Examples

Codex

memex-cli run --backend "codex" --prompt "编码" --stream-format "text"

Claude

memex-cli run --backend "claude" --prompt "设计" --stream-format "text"

Gemini

memex-cli run --backend "gemini" --prompt "UX" --stream-format "text"

Output

Outputs are streamed in the specified format (jsonl or text), allowing real-time monitoring of task progress.

Example JSONL output(multiple jsonl lines)

{"v":1,"type":"assistant.output","ts":"2026-01-08T08:22:20.664800300+00:00","run_id":"a9ba0e5d-9dd5-43a1-8b0f-b1dd11346a2b","action":"\"{}\"","args":null,"output":"{\n  \"mode\": \"command\",\n  \"task_type\": \"general\",\n  \"complexity\": \"simple\",\n  \"backend_hint\": null,\n  \"skill_hint\": null,\n  \"confidence\": 0.92,\n  \"reasoning\": \"简单的文件写入任务，生成10道算术题并写入文件，可用echo或Python命令直接完成\",\n  \"enable_parallel\": false,\n  \"parallel_reasoning\": \"单一文件写入操作，顺序 执行即可\"\n}"}

Example Text output(multiple text lines, any format)

{
  "mode": "backend",
  "task_type": "general",
  "complexity": "simple",
  "backend_hint": "claude",
  "skill_hint": null,
  "confidence": 0.92,
  "reasoning": "生成10道算术题目并写入文件，简单内容生成任务，适合直接LLM处理",
  "enable_parallel": false,
  "parallel_reasoning": "单一文件写入任务，无法分解并行"
}