Is the Agent CLI skill free?

Yes. The skill itself is free from glebis/claude-skills. SkillProof publishes the install command and an independent test verdict at no cost.

What is the Agent CLI SkillProof Score?

9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.

How do I install Agent CLI?

Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.

Can I use Agent CLI with Cursor, Copilot, Gemini CLI, Codex or other AI tools?

The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.

Agent CLI

Convert Python CLIs to agent-consumable --json NDJSON with a shared helper library

by glebis · glebis/claude-skills

Tested · Works

Install — copy the command Download source ↗

Test report

Verdict: Tested · Works
Score: 9.2/10
Tested: Jul 31, 2026
Environment: Claude Code 2.x (agent harness)
Upstream re-checked: Aug 10, 2026 · 6c786e1

Fetched SKILL.md and both referenced files (best-practices.md and evals.json both HTTP 200). Wrote a human-only argparse CLI, then produced a BASELINE conversion (naive inline json.dumps with ad-hoc keys, no timestamps, an as_json param threaded through every function) and a SKILL conversion (copied the reference cli_utils.py verbatim and applied the Step 1-4 patterns). Both ran. Concrete difference: skill output emits a uniform envelope {"event","ts",...} on every line with ISO timestamps and a proper {"event":"error"} on the error path plus exit 1, whereas baseline lines carry inconsistent shapes ("status"/"attempt"/"error") and no timestamps; I confirmed every skill JSON line parses as NDJSON, human mode is auto-preserved via the global flag, and --json appears in --help. I also ran the skill's own provided pytest assertions for json_log and die against cli_utils.py — both passed. No security smells (no curl|sh, base64, or eval).

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

Installs cleanly 5/5
Triggers reliably 5/5
Output vs. baseline 8/10
Docs & honesty 5/5

What Agent CLI does

Adds an agent-friendly --json NDJSON output mode to Python CLI scripts, converting print statements into structured, timestamped JSON events behind a single mode flag, or scaffolds a distributable cli_utils package with tests and CI. Triggers when the user wants machine-readable CLI output for AI agents, asks to add a --json flag, mentions NDJSON or structured CLI output, or wants to build a CLI helper library. Ships a reference cli_utils.py implementation, step-by-step conversion patterns, and a completion checklist.

How to install Agent CLI

git clone --depth 1 https://github.com/glebis/claude-skills.git /tmp/agent-cli-src
mkdir -p ~/.claude/skills
cp -R /tmp/agent-cli-src/agent-cli ~/.claude/skills/agent-cli
# SKILL.md is self-contained Python; no external deps to install.
# References bundled: references/best-practices.md, evals/evals.json
# Package-scaffold mode (Mode B) generates a project that needs pytest>=8 to run its test suite.

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Agent CLI

/agent-cli Convert Python CLIs to agent-consumable --json NDJSON with a shared helper library

It also activates on plain-language prompts like these:

Add a --json flag to this Python script
Convert these print statements to structured JSON
Scaffold a cli_utils package for this project

Frequently asked questions

Is the Agent CLI skill free?: Yes. The skill itself is free from glebis/claude-skills. SkillProof publishes the install command and an independent test verdict at no cost.
Does Agent CLI work with Claude Code?: We tested it with Claude Code 2.x (agent harness) on Jul 31, 2026. Verdict: Tested · Works. Fetched SKILL.md and both referenced files (best-practices.md and evals.json both HTTP 200). Wrote a human-only argparse CLI, then produced a BASELINE conversion (naive inline json.dumps with ad-hoc keys, no timestamps, an as_json param threaded through every function) and a SKILL conversion (copied the reference cli_utils.py verbatim and applied the Step 1-4 patterns). Both ran. Concrete difference: skill output emits a uniform envelope {"event","ts",...} on every line with ISO timestamps and a proper {"event":"error"} on the error path plus exit 1, whereas baseline lines carry inconsistent shapes ("status"/"attempt"/"error") and no timestamps; I confirmed every skill JSON line parses as NDJSON, human mode is auto-preserved via the global flag, and --json appears in --help. I also ran the skill's own provided pytest assertions for json_log and die against cli_utils.py — both passed. No security smells (no curl|sh, base64, or eval).
What is the Agent CLI SkillProof Score?: 9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.
How do I install Agent CLI?: Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Agent CLI with Cursor, Copilot, Gemini CLI, Codex or other AI tools?: The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.