Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/InDate/devharnessWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/plugins/indate/devharness/plugin)<a href="https://agentmods.dev/plugins/indate/devharness/plugin"><img src="https://agentmods.dev/badge/plugins/indate/devharness/plugin/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/plugins/indate/devharness/plugin"><img src="https://agentmods.dev/badge/plugins/indate/devharness/plugin.svg" alt="Reviewed on agentmods" width="80" height="20"></a>Grade A, and why
devharness scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
{
"name": "devharness",
"description": "Run and debug a live app from inside Claude Code: breakpoints and variable inspection in Chrome and Node.js, managed dev servers, and replayable call history so setup is never re-driven by hand.",
"author": {
"name": "InDate",
"url": "https://github.com/InDate"
},
"homepage": "https://github.com/InDate/devharness",
"license": "MIT"
}
What it installs
The manifest is a name and a version. 1 skill, 1 hook, 1 MCP server travel with it, and installing the plugin installs all of them — 147 tokens a session between them. Each is measured on its own page, and each can be installed alone.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 11 lines scan A c44b6231663c
devharness is a plugin published in the GitHub repository InDate/devharness (17 stars, last pushed 7d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other plugins, from other repositories
mochi
A browser companion for AI assistants. Bundled MCP tools for Chrome automation, persistent project memory (Continuum chain), and an in-page hint modal — pair with the Mochi Chrome extension.
cdpilot
Zero-dependency browser automation over raw CDP — 70+ commands, friction ladder, 3-tier stealth, CAPTCHA + press-and-hold solvers, proxy pools, MCP server. No Puppeteer/Playwright/Selenium.
kaboom-workflows
Workflow commands that compose Kaboom MCP's 5 browser telemetry tools into complete, opinionated workflows for debugging, auditing, recording, and interactive development.
groundhog
Search, read and research the web through a real stealth-patched Chrome. Text hidden from human readers is stripped before the model reads it and reported, every source carries a SHA-256 provenance receipt, and the SSRF guard is re-checked after redirects.
rawtrace-mcp
Passive Chromium CDP capture: baselines, incremental changes, and queryable browser history.
tandem
A shared Chrome browser (human + Claude, CDP + Playwright MCP) that learns how each site is navigated: the human drives with the mouse and clears captchas/logins, Claude operates over CDP, and tandem:map persists per-site navigation memory (durable locators, drift, auto-injection).