Cli on Vijay Kodam

model-audit Claude Code SKILL

Sat, 04 Jul 2026 10:00:00 +0300

A Claude Code skill, installable as a plugin, that audits which Claude model handled your sessions. Use it to confirm your coding requests ran on the model you selected (Fable 5, Opus 4.8, and so on) and were not silently served by a different one.

View on GitHub

Why

Claude Code’s built-in commands do not answer “which model produced this response?”. /usage shows plan quota, /cost shows spend, /context shows tokens — none report model provenance.

The ground truth is already on your machine. Claude Code writes every session transcript to ~/.claude/projects/, and every assistant message carries a "model" field with the exact model ID that generated it. model-audit reads that per-message audit trail and rolls it up per session.

Install

As a plugin (recommended):

/plugin marketplace add vijaykodam/model-audit
/plugin install model-audit@model-audit

Or from your shell, outside a session:

claude plugin marketplace add vijaykodam/model-audit
claude plugin install model-audit@model-audit

Run /reload-plugins (or restart Claude Code) to activate.

Use

/model-audit # audit the last 7 days (default)
/model-audit 30 # any number of days
/model-audit all # full session history

Example output

Claude Code sessions active in the last 7 day(s)
LAST ACTIVE PROJECT SESSION MODELS (responses)
2026-07-03 09:47 -vijay-projects-project3 85f5d9a6-ec9f-4dc2-... claude-fable-5(24)
2026-07-02 11:18 -vijay-projects-project1 4a347153-4458-47f8-... claude-fable-5(278)
2026-07-02 07:53 -vijay-projects-project1 1a89419d/agent-ab3a2f9e5caae claude-opus-4-8(21)
2026-06-30 18:06 -vijay-projects-project3 a4b9c607-b1e8-4c92-... claude-opus-4-8(128)
Totals across all sessions:
598 claude-fable-5
504 claude-opus-4-8
51 claude-haiku-4-5-20251001

Model counts are per response, so a session that changed model mid-way shows both models with their proportions. A silent downgrade shows up as an unexpected model count inside a session you expected to be on one model. Subagents appear as separate rows (the SESSION looks like 1a89419d/agent-...) — they legitimately run on different models but never handle the main conversation turns.

Notes

Everything runs locally. It is a small shell script reading files on your own machine; nothing is sent over the internet.
No dependencies beyond standard tools (bash, find, grep, sed, awk). Works on macOS and Linux.
Apache-2.0 licensed.

For full documentation and the source, see the GitHub repo.

OpenClaw Skill for Finland's Helmet Library

Sun, 12 Apr 2026 10:00:00 +0300

An OpenClaw SKILL for the Finnish public library Helmet (Helsinki Metropolitan Area libraries). With it, you can ask your OpenClaw agent from WhatsApp: “Is anything overdue for the kids at the library?” or “Place a hold on AI Engineering by Chip Huyen”.

Demo

Why

I had recently set up OpenClaw and wanted it to simplify my daily life. My kids and I read a lot of books, and we borrow them from Helmet.

Helmet had no CLI or API to manage your library account. It has an open API to search books, but nothing for loans, holds, or fines. So I built the Helmet CLI first, and then a SKILL so that my OpenClaw agent can use it.

What the CLI does

Check loans, and renew one item or all renewable items.
List, place, and cancel holds, with pickup deadline and queue position.
Show fines and a full account summary.
Search the Helmet catalog without a login.
Keep several library cards as profiles, so one command can cover the whole family.

Every command accepts --json, which makes the output easy for an agent to read.

Install

Install the CLI from npm, then save a library card:

npm install -g @helmet-ai/helmet
helmet login

Install the SKILL from ClawHub, or with the skills CLI:

npx skills add vijaykodam/helmet

View on GitHub

url-logger Claude Code Plugin

Mon, 09 Mar 2026 10:00:00 +0200

A Claude Code plugin that logs all WebSearch and WebFetch URLs to .claude/web-access.log, with session IDs and timestamps.

Demo

Why

When you do research or ask Claude Code CLI to search the Internet, it opens many websites. I wanted a way to log all those URLs in a log file.

Claude Code has no setting for this. Claude stores session history in JSONL files, which can contain URLs. But those URLs are mixed with a lot of other information, and they are spread across several session files.

The next option was a hook. But Claude Code’s /hookify cannot write log files; it can only show a warning in the Claude session. This was the perfect use case to build my own hook and publish it as a plugin.

Install

Run these two commands:

# Add the marketplace (one-time)
claude plugin marketplace add vijaykodam/url-logger

# Install the plugin
claude plugin install url-logger

Restart Claude Code CLI after the install. The plugin starts to work only after the restart.

Log format

[2026-03-07T14:23:01Z] session=abc123 tool=WebSearch query="claude code plugins"
[2026-03-07T14:23:05Z] session=abc123 tool=WebFetch url="https://example.com/docs"

By default, the plugin writes the log to .claude/web-access.log in your working directory. To use one log for all projects, set URL_LOGGER_PATH:

export URL_LOGGER_PATH=~/.claude/web-access.log

The plugin uses a Claude Code PreToolUse hook. It works only in Claude Code CLI, not in Claude.ai or the Claude Desktop app.

View on GitHub