Agent skill

CLI-Anything for Codex

by HKUDS in HKUDS/CLI-Anything

Lets Codex build, refine, test, validate and list CLI-Anything harnesses for GUI applications or source repositories, following the project's full methodology.

Apache-2.0Auto-check passedDevelopment

Install CLI-Anything for Codex

skills CLI
$ npx skills add HKUDS/CLI-Anything --skill cli-anything -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install HKUDS/CLI-Anything cli-anything --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/HKUDS/CLI-Anything.git skills-src && mkdir -p .claude/skills && cp -r skills-src/codex-skill .claude/skills/cli-anything && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
cli-anything
GitHub stars
52k
Token cost
~1.5k tokens
SKILL.md length
547 words
Files
5 (incl. scripts)
Skills in repo
78
Repo updated
First seen
Licence
Apache-2.0

At a glance

Lets Codex build, refine, test, validate and list CLI-Anything harnesses for GUI applications or source repositories, following the project's full methodology.

  • Works in 6 steps: Read references/HARNESS.md relative to… → Read the matching mode specification… → Read referenced files under… → …
  • Building a CLI-Anything harness for a GUI application
  • SKILL.md covers Read the Full Methodology First, Resource Map, Inputs and Modes, plus 2 more sections
  • Runs Shell and PowerShell scripts from its folder

What it does

This is the Codex adaptation of the CLI-Anything builder. Before any work it has the agent read the full methodology in references/HARNESS.md, then the specification for the mode in use under references/commands/ (build, refine, test, validate or list), and the guides under references/guides/ only when they apply. If those files are missing it falls back to a CLI-Anything repository checkout, then to cloning the project, and only as a last resort to condensed rules.

The installer vendors the canonical resources into the skill so a normal Codex install is self-contained, and document paths are remapped to the installed locations. Bundled scripts include repl_skin.py and preview_bundle.py, which are copied into generated harnesses, and skill_generator.py with a SKILL.md template for producing CLI skills. Install scripts for shell and PowerShell come with an install test. The generated Python harness format is unchanged.

When your agent uses it

  • Building a CLI-Anything harness for a GUI application
  • Refining or testing an existing generated harness
  • Validating a harness against the methodology checklist
  • Listing installed or generated harnesses

Example prompts

  • “Build a CLI-Anything harness for the open-source app checked out in ./apps/sketchpad.”
  • “Run the test mode on the harness we generated yesterday and report the failures.”
  • “List the CLI-Anything harnesses installed on this machine.”
  • “Validate the harness in ./harnesses/sketchpad against the checklist.”

Requirements

  • Python, for the generated harnesses and helper scripts
  • Network access to GitHub if the bundled resources are missing

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

  1. Read references/HARNESS.md relative to this skill directory.
  2. Read the matching mode specification under references/commands/.
  3. Read referenced files under references/guides/ only when they apply to the target.
  4. If the installed resources are unavailable and this skill is being used from a
  5. If neither local source is available, clone or download
  6. Only if all full-methodology sources are unavailable, follow the condensed rules below.

What it can do on your machine

Read from SKILL.md and the folder at commit 34f5195. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 2 files in scripts/ (Shell and PowerShell), which the agent can run.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

CLI-Anything for Codex loads about 1.5k tokens when it runs. Until then it costs about 64 tokens; SKILL.md has 547 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~64
When it runs · the whole SKILL.md, loaded when a task matches
~1.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from HKUDS/CLI-Anything at commit 34f5195, republished under its Apache-2.0 licence (© HKUDS). 547 words, ~1,502 tokens.

Download SKILL.mdSave it as .claude/skills/cli-anything/SKILL.md (or your agent's skills folder). This skill also uses 4 other files; get the full folder from GitHub.
name
cli-anything
description
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.

CLI-Anything for Codex

Use this skill when the user wants Codex to act like the CLI-Anything builder.

Read the Full Methodology First

Before implementation, use the full methodology source of truth:

  1. Read references/HARNESS.md relative to this skill directory.
  2. Read the matching mode specification under references/commands/.
  3. Read referenced files under references/guides/ only when they apply to the target.
  4. If the installed resources are unavailable and this skill is being used from a CLI-Anything repository checkout, read ../cli-anything-plugin/HARNESS.md and the resources around it.
  5. If neither local source is available, clone or download https://github.com/HKUDS/CLI-Anything and use cli-anything-plugin/HARNESS.md.
  6. Only if all full-methodology sources are unavailable, follow the condensed rules below.

The installer vendors the canonical resources from cli-anything-plugin/ into this skill so a normal Codex installation is self-contained.

Resource Map

PathPurpose
references/HARNESS.mdComplete methodology and quality rules
references/commands/cli-anything.mdBuild-mode specification
references/commands/refine.mdRefine-mode specification
references/commands/test.mdTest-mode specification
references/commands/validate.mdValidation checklist
references/commands/list.mdInstalled/generated harness discovery
references/guides/On-demand implementation guidance
scripts/repl_skin.pyCopy into generated harnesses as utils/repl_skin.py
scripts/preview_bundle.pyCopy into preview-capable harnesses as utils/preview_bundle.py
scripts/skill_generator.pyGenerate canonical and packaged CLI skills
scripts/templates/SKILL.md.templateSkill generation template used by skill_generator.py
references/docs/PREVIEW_PROTOCOL.mdShared preview bundle protocol

When reading vendored documents, apply these path remappings instead of resolving plugin paths against the current working directory:

Document referenceInstalled skill path
guides/...references/guides/...
cli-anything-plugin/repl_skin.pyscripts/repl_skin.py
cli-anything-plugin/preview_bundle.pyscripts/preview_bundle.py
cli-anything-plugin/skill_generator.pyscripts/skill_generator.py
templates/SKILL.md.templatescripts/templates/SKILL.md.template
docs/PREVIEW_PROTOCOL.mdreferences/docs/PREVIEW_PROTOCOL.md

Inputs

Accept either:

  • A local source path such as ./gimp or /path/to/software
  • A GitHub repository URL

Derive the software name from the local directory name after cloning if needed.

Modes

Build

Read references/commands/cli-anything.md. Build a complete harness through source analysis, architecture design, implementation, test planning, test implementation, test documentation, CLI-specific SKILL.md generation, and local installation.

Refine

Read references/commands/refine.md. Inventory current commands and tests, compare them against the target software, then incrementally add high-impact capabilities without removing existing commands unless the user requests a breaking change.

Test

Read references/commands/test.md. Run the harness tests against the real backend, verify installed-command subprocess coverage, and update TEST.md only with passing results.

Show full SKILL.md (212 more words)Show less
Validate

Read references/commands/validate.md. Validate the harness against the complete directory, implementation, test, documentation, packaging, and code-quality checklist.

List

Read references/commands/list.md. Discover installed and generated CLI-Anything tools and support human-readable or JSON output as requested.

Condensed Fallback Rules

Use these only when the full methodology cannot be retrieved.

Harness Structure

Produce:

text
<software>/
└── agent-harness/
    ├── <SOFTWARE>.md
    ├── setup.py
    └── cli_anything/
        └── <software>/
            ├── README.md
            ├── __init__.py
            ├── __main__.py
            ├── <software>_cli.py
            ├── core/
            ├── utils/
            └── tests/
Required Behavior
  • Prefer the real software backend over reimplementation.
  • Provide one-shot Click subcommands and default REPL mode.
  • Support --json machine-readable output.
  • Add session state with undo/redo where the target supports it.
  • Auto-save one-shot session mutations and support --dry-run.
  • Use locked session-file writes to avoid concurrent JSON corruption.
  • Copy and use the unified ReplSkin.
  • Add truthful preview commands for software with meaningful visual or inspection state.
  • Generate both canonical and packaged CLI-specific SKILL.md files.
Testing
  • Write TEST.md before test code.
  • Keep test_core.py for unit coverage.
  • Keep test_full_e2e.py for real-file workflows and real backend validation.
  • Verify rendered/exported output programmatically, not only process exit codes.
  • Test the installed cli-anything-<software> command via _resolve_cli().
  • Run release validation with CLI_ANYTHING_FORCE_INSTALLED=1.
Packaging
  • Use find_namespace_packages(include=["cli_anything.*"]).
  • Keep cli_anything/ as a namespace package without a top-level __init__.py.
  • Expose cli-anything-<software> through console_scripts.
  • Include the packaged CLI-specific skills/SKILL.md.

Output Expectations

When reporting progress or final results, include:

  • target software and source path
  • files added or changed
  • validation commands run
  • open risks or backend limitations

© HKUDS, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 4 other files (scripts) in codex-skill of HKUDS/CLI-Anything.

  • SKILL.md
  • agents/openai.yaml
  • scripts/install.ps1
  • scripts/install.sh
  • tests/test_install.sh

Open the folder on GitHubat commit 34f5195

Compare with similar skills

CLI-Anything for Codex next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

CLI-Anything for Codex compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
CLI-Anything for Codex this skillHKUDS/CLI-Anything52k—~1.5kAutomated safety check: PassApache-2.0
Skill Creatorluongnv89/asm953—~5.3kAutomated safety check: PassMIT
Skill CreatorIgorWarzocha/Opencode-Workflows122—~2.9kAutomated safety check: PassApache-2.0
Mirage VFS Adapter Authoringstrukto-ai/mirage3.7k—~2.4kAutomated safety check: PassApache-2.0
Bmad Builderaj-geddes/claude-code-bmad-skills487—~1.9kAutomated safety check: NotesCustom licence
Memory Starter Kitmage0535/hermes-memory-installer209—~363Automated safety check: PassMIT

Similar skills

  • Skill Creator

    luongnv89/asm

    Create a skill or bring an existing one up to the same standard (validate + asm eval fix loop); run evals, tune triggering.

    953 GitHub stars~5.3k tokensUpdated yesterday
    Agent WorkflowsAuto-check passed
  • Skill Creator

    IgorWarzocha/Opencode-Workflows

    Architect and refine OpenCode skills to extend agent capabilities.

    122 GitHub stars~2.9k tokensUpdated 8 mo ago
    Agent WorkflowsAuto-check passed
  • Builds or extends a custom Mirage virtual filesystem adapter for an API, database, object store or app data, with a working mount configuration and filesystem tests.

    3.7k GitHub stars~2.4k tokensUpdated today
    DevelopmentAuto-check passed
  • Bmad Builder

    aj-geddes/claude-code-bmad-skills

    Meta-skill for scaffolding and validating custom PLANNING/ORCHESTRATION skills within the BMAD Planning & Orchestrator plugin.

    487 GitHub stars~1.9k tokensUpdated 3 mo ago
    DevelopmentAuto-check: notes
  • Memory Starter Kit

    mage0535/hermes-memory-installer

    Historical starter note for the memory sidecar stack. An agent skill from mage0535/hermes-memory-installer.

    209 GitHub stars~363 tokensUpdated 1 mo ago
    DevelopmentAuto-check passed
  • Create Skill

    dotnet/skills

    Official

    Scaffolds new agent skills for the dotnet/skills repository.

    5.6k GitHub starsUsed in 1 repo~2.7k tokens
    DevelopmentAuto-check passed

More from HKUDS/CLI-Anything

All 78 skills in this repo
  • CLI-Anything for Reasonix

    HKUDS/CLI-Anything

    Adapts the CLI-Anything methodology so Reasonix can build, refine, test and validate a command-line harness for an existing GUI application or source repository.

    52k GitHub stars~1.8k tokensUpdated 15 days ago
    Auto-check passed
  • Guides Hermes Agent through building, refining and testing a CLI-Anything harness that wraps a GUI application or source repository in a scriptable command line.

    52k GitHub stars~1.1k tokensUpdated 15 days ago
    Auto-check passed
  • Builds, refines, tests or validates a CLI-Anything harness that wraps a GUI application or source repository as a stateful Click CLI with JSON output and a REPL mode.

    52k GitHub stars~839 tokensUpdated 15 days ago
    Auto-check passed
  • Krita CLI Harness

    HKUDS/CLI-Anything

    Drives Krita from the command line to create projects, add layers, apply filters, resize the canvas and export images, with JSON output for agents.

    52k GitHub stars~931 tokensUpdated 15 days ago
    Auto-check passed
  • MacroCLI GUI Macros

    HKUDS/CLI-Anything

    Defines, lists, inspects and runs parameterized GUI macros from the command line, with the runtime choosing the execution backend.

    52k GitHub stars~1.4k tokensUpdated 15 days ago
    Auto-check passed
  • Capability-based matrix for video production in which the agent picks a provider per capability, from storyboard and sound design to captions, render checks and packaging.

    52k GitHub stars~12k tokensUpdated 15 days ago
    Auto-check passed

Works with

Questions about CLI-Anything for Codex

What does CLI-Anything for Codex do?

Lets Codex build, refine, test, validate and list CLI-Anything harnesses for GUI applications or source repositories, following the project's full methodology. This is the Codex adaptation of the CLI-Anything builder.md, then the specification for the mode in use under references/commands/ (build, refine, test, validate or list), and the guides under references/guides/ only when they apply.

When should I use CLI-Anything for Codex?

CLI-Anything for Codex fits situations like: building a CLI-Anything harness for a GUI application; refining or testing an existing generated harness; validating a harness against the methodology checklist; listing installed or generated harnesses.

How do I install CLI-Anything for Codex in Claude Code?

Run `npx skills add HKUDS/CLI-Anything --skill cli-anything -a claude-code`. Or copy the skill folder (codex-skill in HKUDS/CLI-Anything) into .claude/skills/cli-anything in your project. Claude Code loads it when a task matches its description.

How do I install CLI-Anything for Codex in Codex?

Run `npx skills add HKUDS/CLI-Anything --skill cli-anything -a codex`. Or copy the skill folder (codex-skill in HKUDS/CLI-Anything) into .agents/skills/cli-anything in your project. Codex loads it when a task matches its description.

Can I use CLI-Anything for Codex in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add HKUDS/CLI-Anything --skill cli-anything -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cli-anything, .gemini/skills/cli-anything, .github/skills/cli-anything and .opencode/skills/cli-anything in your project.

What does CLI-Anything for Codex need to run?

Going by SKILL.md and its folder, CLI-Anything for Codex needs a shell and PowerShell for the scripts in its folder. Our summary lists: Python, for the generated harnesses and helper scripts; Network access to GitHub if the bundled resources are missing.

Does CLI-Anything for Codex access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is CLI-Anything for Codex safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does CLI-Anything for Codex use?

CLI-Anything for Codex is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does CLI-Anything for Codex use?

About 1.5k tokens (SKILL.md is roughly 6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to CLI-Anything for Codex?

Skills that share tags, products or a category with CLI-Anything for Codex: Skill Creator (luongnv89/asm, 953 stars), Skill Creator (IgorWarzocha/Opencode-Workflows, 122 stars), Mirage VFS Adapter Authoring (strukto-ai/mirage, 3.7k stars) and Bmad Builder (aj-geddes/claude-code-bmad-skills, 487 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains CLI-Anything for Codex?

HKUDS (a GitHub organization) maintains it in HKUDS/CLI-Anything, which has 51,644 GitHub stars. The repository holds 78 skills in this directory. The repository was last updated on September 22, 2026.

Source: HKUDS/CLI-Anything on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.