JJ Englert had Claude Code build a "Search Radar" mod that counts empty searches and unneeded file reads in each answer and writes a shortcut doc in one click
On 2026-10-19 JJ Englert (@JJEnglert) pointed out that every time Claude Code looks for something in your project it reads files, searches and runs bash; some lookups come back empty and some read files it never needed, which is time and tokens you never see. With one prompt he had Claude build a mod (Claude Code's early-access function-hooks extensions, which can add UI above the prompt) called Search Radar: it scores the wasted lookups for each question and, when a search was wasteful, offers a button that has Claude write a shortcut note into your docs so the next agent goes straight there. He tried it with two questions on his own repo; the second showed his folder layout needed work. We haven't run it, and his public claude-mods repo doesn't contain Search Radar yet, so for now you generate it from his prompt.
Visit site Claude Code Mods (function hooks) ↗- Pricing
- Paid
- Region
- Global
- Where it's from
- Overseas
- Added
- Oct 24, 2026
- Version we used
- Post dated 2026-10-19; JJenglert1/claude-mods main @ f6bd7fc (2026-10-13, no Search Radar); mods built against Claude Code 2.1.290 per that repo
- Categories
- CodingProductivity
- Original author
- JJ Englert @JJEnglert ↗
- Our test
- Not tested by us
- Cost / time
- Requires a Claude plan or API access that includes Claude Code; building the mod and each shortcut-doc click use extra tokens.
The task
Find out where a coding agent wastes effort finding things in your repo (empty searches, files read for nothing), write the route back into your docs so later agents go straight there, and use that as a test of how navigable your folder and doc structure is.
Steps
- Prerequisite: your Claude Code must support Mods (the author describes them as small upgrades inside Claude Code that you make by describing what you want; no code). His claude-mods repo notes the function-hooks API behind Mods is early access and may change between releases, and that his mods were built and checked against Claude Code 2.1.290.
- Paste the prompt below (his exact wording) into Claude Code to have it build Search Radar: a live meter above the prompt that scores how efficiently each of your questions was answered.
- Start a new chat afterwards so the mod loads (his instruction for his own mods too).
- Ask questions as usual. The radar watches every lookup Claude makes while answering; an empty result or a file read it didn't need counts as waste.
- When it flags waste, click the button: Claude writes a short shortcut note in your docs so the next agent knows exactly where to go.
- Use it as a gauge for your structure. Asking "how does the repo judge whether work is good quality?" found the answer fast but read one unneeded skill (patched in one click); asking it to bring back a hidden mod produced two repeated reads and a pile of unnecessary searching through hook files, which told him his folder setup needed work.
- If it earns its keep, enable it at user level; he runs it in every project he opens.
Prompt
Build me a mod called Search Radar. It’s a live meter shown above the prompt. For each question I type, it scores how effectively you found what you needed and offers a button to write a shortcut doc when the search was wasteful, so the agent knows how to get there faster next time.
What the author says
He says this is the first time he can actually see how well his agents move through his file and folder system. In two tests on his own repo, one answer read just one unneeded skill (shortcut patched in one click), and the other exposed repeated reads and needless hook-file searching, showing the folder layout needed work. He now keeps it on at user level. The post includes a roughly 4-minute demo video. He gives no figures for tokens or time saved.
Our test
We haven't built or run Search Radar; results are the author's claims and demo. We only cloned and checked the claude-mods repo he links (main @ f6bd7fc): none of its eight mods is Search Radar.
How to get in
The original is an English X post with a demo video; the claude-mods repo is on GitHub.
Who it suits
People with large repos full of docs and skills who feel Claude Code spends ages hunting; anyone restructuring CLAUDE.md or docs who wants to see how the agent actually navigates.
Pitfalls
- He says "or grab mine, all my Mods are in the link below", but his public JJenglert1/claude-mods repo (latest commit 2026-10-13, before this post) has eight mods and no Search Radar. There is a similarly named Usage Radar, which tracks plan usage; don't install the wrong one. For now you have to generate it from the prompt.
- Mods rely on the early-access function-hooks API, which his repo says may change between releases; if your Claude Code version doesn't support it, this won't work.
- "Waste" is judged by a Claude-built mod, so whether a read was truly unneeded is partly subjective; he shows two examples and doesn't quantify savings.
- The button makes Claude write into your docs. Review where and what it writes before committing so your docs don't fill up with scattered shortcut notes.
- The prompt is a natural-language spec, so the mod Claude builds for you may differ from his, and results may not match the demo.
More in this category
- A demo video per docs page that flags itself when the UI changesVideoScreenplayClaude CodeTool free; uses your agent's quota◐ Partly testedOverseas
- AI backend code that guards against races, leaks and bad migrationsCodingbackendskillsClaude CodePostgreSQLFree; about 2× tokens per task◐ Partly testedOverseas
- An afternoon of AI work, a 3D action game you play in the browserCodingClaude CodeTripoJimengSkill free; Claude quota plus asset tools◐ Partly testedChina
- Keep a long-running AI-built project from turning into a messCodingClaudeCodexoil-delegateNeeds Claude and Codex plans◐ Partly testedChina
- Use a web app once and get an AI write-up of how it worksResearchREAClaude CodeChromeFree tools; needs Claude or Codex◐ Partly testedOverseas