cli-anything

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.

1.55x

Quality

62%

Does it follow best practices?

Impact

95%

1.55x

Average score across 3 eval scenarios

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./codex-skill/SKILL.md

Quality

Content

57%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The SKILL.md is well-organized and token-efficient as an overview, but its core instructional strategy delegates to bundled reference files that are not actually present, leaving only the condensed fallback rules as concrete guidance. This weakens actionability, workflow validation, and the progressive-disclosure structure that depends on those files.

Suggestions

Vendor the referenced files into the bundle (references/HARNESS.md, references/commands/*.md, scripts/repl_skin.py, etc.) so the resource map points to real paths; a normal install currently resolves none of them.

Add an inline validate→fix→retry feedback loop for the Build/Test workflows (e.g., run tests → on failure review error → fix → re-run) rather than delegating all checkpoints to the missing validate.md.

Inline one concrete, copy-pasteable example per mode (e.g., a minimal Click subcommand skeleton or a namespace-package setup.py snippet) so the skill is actionable even before the referenced files are read.

Dimension	Reasoning	Score
Conciseness	The body is a lean overview that assumes Claude's competence (no explanations of CLI/Click basics) and uses condensed bullet rules, though the Resource Map and path-remapping tables redundantly enumerate the same files and the six-step retrieval cascade could be trimmed.	4 / 5
Actionability	The primary mode sections only delegate ('Read references/commands/X.md') to files that are not actually bundled, while the concrete executable details (namespace-package include pattern, CLI_ANYTHING_FORCE_INSTALLED=1, _resolve_cli()) live only in the condensed fallback and are fragmented.	3 / 5
Workflow Clarity	The Build sequence is listed as a flat comma-list (source analysis → architecture → implementation → ...) and validation is delegated to a separate Validate mode with no inline validate→fix→retry feedback loop, which caps a batch/code-generating skill at 3.	3 / 5
Progressive Disclosure	Structure is well-intended with a clear resource map and one-level-deep references, but the actual bundle is missing every referenced file (no references/ dir; none of repl_skin.py, preview_bundle.py, skill_generator.py, HARNESS.md, commands/*.md exist), so the disclosure is non-functional in a normal install and relies on the GitHub-clone fallback.	3 / 5
	Total	13 / 20 Passed

Description

67%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description gives a clear, third-person trigger with five concrete actions and an explicit 'Use when...' clause, scoring well on completeness and distinctiveness. It is held back by reliance on the proprietary 'CLI-Anything' term, which limits natural trigger-term quality.

Suggestions

Add natural synonyms a user might say without knowing the project name (e.g., 'CLI wrapper', 'command-line interface harness', 'REPL CLI generator') to broaden trigger coverage.

Briefly define 'CLI-Anything harness' in plain terms within the description so the 'what' is self-contained rather than depending on external methodology knowledge.

Dimension	Reasoning	Score
Specificity	Names five concrete verbs ('build, refine, test, validate, or list') scoped to 'CLI-Anything harnesses for GUI applications or source repositories', giving several specific actions with only minor gaps in the abstractness of 'harnesses'.	4 / 5
Completeness	Both 'what' (adapts the CLI-Anything methodology to Codex) and 'when' (explicit 'Use when the user wants Codex to build, refine, test, validate, or list...') are present, though the 'what' leans on external jargon rather than being fully self-contained.	4 / 5
Trigger Term Quality	The action verbs are natural, but the trigger relies heavily on the project-specific jargon 'CLI-Anything harnesses' and lacks common synonyms or file extensions users might naturally say.	3 / 5
Distinctiveness Conflict Risk	The 'CLI-Anything harnesses' niche is distinctive with minimal overlap risk, though the generic build/test/validate verbs could marginally overlap with related tooling skills.	4 / 5
	Total	15 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
referenced_paths_exist	Referenced path issues: 26 missing, 14 deeper-than-1-level	Warning

	Total	15 / 16 Passed

Repository: HKUDS/CLI-Anything
Path: codex-skill/SKILL.md
Commit: 810c18b

Reviewed: 8 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.