本文へ移動
cccskills
無料GitHub で公開

cli-anything-macrocli

Use when the agent wants to define, list, inspect, or execute GUI macros via the MacroCLI CLI. Macros are parameterized, CLI-callable workflows — the agent invokes `macro run <name>` and the system handles backend routing (plugin, file transform, accessibility, compiled GUI replay).

インストール方法を見る

含まれるファイル(1)

  • SKILL.md5.6 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

MacroCLI CLI

What It Is

The MacroCLI converts valuable GUI workflows into parameterized, CLI-callable macros. Agents never touch the GUI directly — they call macros through this stable CLI, and the runtime routes execution to the best available backend (native plugin/API, file transformation, semantic UI control, or precompiled GUI macro replay).

Installation

cd macrocli/agent-harness
pip install -e .

Requirements: Python 3.10+, PyYAML, click, prompt-toolkit.

Quick Start (for agents)

# 1. See what macros are available
cli-anything-macrocli macro list --json

# 2. Inspect a macro's parameters
cli-anything-macrocli macro info export_file --json

# 3. Dry-run to check params without side effects
cli-anything-macrocli --dry-run macro run export_file \
    --param output=/tmp/test.txt --json

# 4. Execute a macro
cli-anything-macrocli macro run export_file \
    --param output=/tmp/result.txt --json

# 5. See what backends are available
cli-anything-macrocli backends --json

Command Reference

Global Flags

FlagDescription
--jsonMachine-readable JSON output on stdout
--dry-runSimulate all steps, skip side effects
--session-id <id>Resume or create a named session

macro group

CommandDescription
macro listList all available macros
macro info <name>Show macro schema (parameters, steps, conditions)
macro run <name> --param k=vExecute a macro
macro dry-run <name> --param k=vSimulate without side effects
macro validate [name]Structural validation
macro define <name>Scaffold a new macro YAML

session group

CommandDescription
session statusShow session statistics
session historyShow recent run history
session savePersist session to disk
session listList all saved sessions

backends

cli-anything-macrocli backends --json
# Shows: native_api, file_transform, semantic_ui, gui_macro, recovery
# and whether each is available in the current environment.

Macro Parameters

Pass parameters with --param key=value. Repeat for multiple:

cli-anything-macrocli macro run transform_json \
    --param file=/path/to/data.json \
    --param key=settings.theme \
    --param value=dark \
    --json

Output Format (--json)

All commands output JSON when --json is set:

{
  "success": true,
  "macro_name": "export_file",
  "output": {
    "exported_file": "/tmp/result.txt"
  },
  "error": "",
  "telemetry": {
    "duration_ms": 312,
    "steps_total": 2,
    "steps_run": 2,
    "backends_used": ["native_api"],
    "dry_run": false
  }
}

On failure ("success": false), read the "error" field for the reason. Exit code is 1 on failure.

Execution Backends

Backends are selected automatically based on the macro step definition:

BackendTriggered byUse case
native_apibackend: native_apiSubprocess / shell command
file_transformbackend: file_transformXML, JSON, text file editing
semantic_uibackend: semantic_uiAccessibility / keyboard shortcuts
gui_macrobackend: gui_macroPrecompiled coordinate replay
recoverybackend: recoveryRetry / fallback orchestration

Writing Macros

Macros are YAML files in cli_anything/macrocli/macro_definitions/. Scaffold one with:

cli-anything-macrocli macro define my_macro --output \
    cli_anything/macrocli/macro_definitions/examples/my_macro.yaml

Minimal schema:

name: my_macro
version: "1.0"
description: What this macro does.

parameters:
  output:
    type: string
    required: true
    description: Where to write results.
    example: /tmp/result.txt

preconditions:
  - file_exists: /path/to/input

steps:
  - id: step1
    backend: native_api
    action: run_command
    params:
      command: [my-app, --export, "${output}"]
    timeout_ms: 30000
    on_failure: fail  # or: skip, continue

postconditions:
  - file_exists: ${output}
  - file_size_gt: [${output}, 100]

outputs:
  - name: result_file
    path: ${output}

agent_hints:
  danger_level: safe  # safe | moderate | dangerous
  side_effects: [creates_file]
  reversible: true

Agent Usage Rules

  1. Always use --json for programmatic output.
  2. Use --dry-run to validate params before executing side-effectful macros.
  3. Check success field — do not assume success from exit code alone.
  4. Read error field when success is false — it explains what failed.
  5. Use macro info <name> to discover params before calling macro run.
  6. Use absolute paths for all file parameters.

Example Workflow

# Step 1: What's available?
cli-anything-macrocli macro list --json

# Step 2: What params does transform_json need?
cli-anything-macrocli macro info transform_json --json

# Step 3: Test safely
cli-anything-macrocli --dry-run macro run transform_json \
    --param file=/tmp/config.json \
    --param key=theme \
    --param value=dark --json

# Step 4: Execute for real
cli-anything-macrocli macro run transform_json \
    --param file=/tmp/config.json \
    --param key=theme \
    --param value=dark --json

Version

1.0.0

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Use when the user wants Codex to build, refine, test, validate, or list CLI-Anything harnesses for GUI applications or source repositories. Adapts the full CLI-Anything methodology to Codex without changing the generated Python harness format.

日本語の概要は準備中です。原文の説明を表示しています。

HKUDS/CLI-Anything5.2万2026年9月22日 更新

Use when the user wants OpenClaw to build, refine, test, or validate a CLI-Anything harness for a GUI application or source repository. Adapts the CLI-Anything methodology to OpenClaw without changing the generated Python harness format.

日本語の概要は準備中です。原文の説明を表示しています。

HKUDS/CLI-Anything5.2万2026年9月22日 更新

Use when the user wants Reasonix to build, refine, test, or validate a CLI-Anything harness for a GUI application or source repository. Adapts the CLI-Anything methodology to Reasonix without changing the generated Python harness format.

日本語の概要は準備中です。原文の説明を表示しています。

HKUDS/CLI-Anything5.2万2026年9月22日 更新

3MF mesh geometry editor — detect and resize cylindrical holes, repair meshes, compare 3D printing files. Works with BambuStudio and PrusaSlicer 3MF files.

日本語の概要は準備中です。原文の説明を表示しています。

HKUDS/CLI-Anything5.2万2026年9月22日 更新

Command-line interface for AdGuard Home - Network-wide ad blocking and DNS management via AdGuard Home REST API. Designed for AI agents and power users who need to manage filtering, DNS rewrites, clients, DHCP, and query logs without a GUI.

日本語の概要は準備中です。原文の説明を表示しています。

HKUDS/CLI-Anything5.2万2026年9月22日 更新

Command-line interface for Anygen - A stateful command-line interface for AnyGen OpenAPI — generate professional slides, documents, webs...

日本語の概要は準備中です。原文の説明を表示しています。

HKUDS/CLI-Anything5.2万2026年9月22日 更新

HKUDS のスキルをすべて見る

このスキルの問題を報告する