Claude Code Mods

KevinYu1580/jcGearbox

Picks the model once per session and the effort every turn, using TypeSafe's Jev to judge how hard the task is.

karimdjaafar/jc-hud-mod

JC HUD — mod Claude Code : barre d'instruments Jasmine Conseil (contexte, forfait, modèle, horloges)

gblachstein/claude-mods

A dancing GIF at the bottom left that moves while Claude is working and freezes when it stops

jetsongdev/jet-router

Experimental Claude Code effort router: fake and opt-in Jev shadow

gecm0/jev-mod

Adds $.jev: typed Choice, Score and Noul judgments from TypeSafe's Jev, with a timeout and validated answers, for other mods to call.

yinjs/claude-jev

Jev answers clear permission prompts, flags prompt injection in fetched content, checks done claims, nudges large or vague requests, and routes effort and models without trading away the prompt cache.

HAR5HA-7663/jev-compact

Context compaction that prunes instead of summarising: rules drop exploration output and hook noise locally, Jev ranks the rest against a budget, a note says what was removed, and removed outputs are archived for the brain. Falls back to the built-in summary when it cannot do better.

HAR5HA-7663/jev-compact

Companion to jev-compact: watches context use after every turn and starts a compaction at the configured percentage. Separate plugin on purpose — the engine skips a plugin's own session.compact hook when that plugin raises the compaction, so the trigger must come from outside jev-compact for the pruning to run.

klefix/claude-code-mods

Asks Jev whether a prompt leaves a choice open and whether each edit stays within what the user approved

kuhku/kuhaku-plugins

Annotates Claude Code's AskUserQuestion dialog with TypeSafe Jev's probabilities and a recommended option (labels in Japanese).

m-esm/jev-model-router

TypeSafe Jev scores each prompt and switches the session's model and effort to match, per turn.

Tolsee/dotfiles

Sets the model of every subagent spawn (agent.spawn). The agent type sets the baseline: agents whose definition pins a model keep it, general-purpose, claude and Plan get sonnet, and TypeSafe's Jev may move those down to haiku, or up to opus when the task is risky (production, money, irreversible data). Sonnet is the ceiling otherwise; a description starting with [opus] goes to opus. No key or no verdict in time leaves the baseline, never the parent's Opus.

Syptern/jev-prompt-grader

Grades each prompt you send with TypeSafe's Jev model and shows what is missing above the prompt box.

Xvectorio/xvector-claude-mods

Per-turn model routing: TypeSafe Jev picks Haiku, Sonnet or Opus for each prompt (port of gargpratyush/jev-router).

staticdreams/jev-semaphore

Jev Semaphore for Claude Code: Jev / Kev model routing for subagents, with a live allocation pane, agent graph and context-exchange band

AnatoliyShk/bulgolingo

Keeps the skill listing out of the context window and lets TypeSafe's Jev, a System One decision model, pick at most one skill per prompt from the skills' descriptions, the way TypeSafe's skill-suggestion cookbook does: one request ranks every skill and asks whether the prompt needs a skill at all, a second re-reads the top three with their SKILL.md and can reject them all. The winner's SKILL.md is attached to the prompt, so the skills can be hidden from the model altogether (/jev-skill-suggestion:setup makes them user-invocable-only and /context counts them at 0). Falls back to the engine's built-in classifier when no key is set.

sohryuu101/jev-vault-gate

Uses Jev to spot durable knowledge in a turn and auto-capture it into ~/vault/raw/ for /compile.

deadczarvc-labs/jev-watch

Logs what fast-jev-compaction does (outcomes, TypeSafe requests, compactions, hook failures) to JSONL files.

jevcomp/jevcomp

Jev decides what Claude Code keeps when it compacts, instead of a model-written summary.

Fyzu/claude-atlassian-mode

Bitbucket review queue and your open PRs in a pane; the current branch's PR, CI and Jira issue in a bar above the prompt

meganemura/jira-ticket-pane

A pane beside the transcript that fetches one Jira issue by key and lets you attach it to your next prompt.

polatmurat/temsa-claude-mods

Tracks the Jira ticket you work on, offers to open one at commit time and closes it out, writing to Jira only after you approve each step.

Danh295/dannys-claude-mods

A notes pane: jot text during a session, keep it across sessions, arrow to a note and press Enter to paste it into the prompt.

hptukurou0101-design/claude-mods-jp

スピナーや操作ヒントなど Claude Code の英語表示を日本語にする

GururOzatmaca/claude-mods

A side pane listing every prompt in the session: see where you are, click to jump back

orvn/justchat-claude-mod

A /justchat mode where Claude supresses actions like writing code things through instead of writing code in favor of just chatting until told otherwise

holdennguyen/claude-code-setup

A live, interactive k9s terminal in a pane (k9s runs under a private tmux server, so kube auth and context come from the shell that launched Claude Code), plus two tools so the model can read the k9s screen and send it keys. While k9s runs, Bash calls that invoke kubectl are refused and the model is pointed at those tools; kubectl works as usual otherwise. One server per Claude Code process, a watchdog and a start-up sweep keep orphans from surviving, and an idle stop keeps k9s from polling the cluster unseen.

Aurumdev952/claude-code-kamehameha

Context usage as a pixel-art showdown: a hero laughs off a kamehameha that grows with the context window, strains harder and harder, and at 95% it's GAME OVER, please compact

4DvAnCeBoY/kane-qe

KaneAI (kane-cli v16) cockpit for QE engineers in Claude Code: run browser tests in plain English, watch them live, see who owns a failure from the evidence, run suites on the cloud grid and devices, and turn requirements into proven coverage

LambdaTest/kane-claude-plugin

Watches kane-cli (KaneAI) runs in Claude Code: a band above the prompt with the kane mascot, a pane for runs, steps, failures and assurance coverage, and a nudge to test what Claude just changed

aieo-product/kaneed-dungeon-claudemod

カニード が Claude の作業中にダンジョンを自動探索する Claude Mod。回復は Lv アップとテスト成功、あとはマップで見つけたショップだけ。倒れると Lv 1 からやり直し。会話ログのパス・#N・URL もクリックで開ける。Needs function hooks (CLAUDE_CODE_ENABLE_FUNCTION_HOOKS=1) and an interactive terminal.

stelmach-systems/claude-mods

Skrzynka nieprzeczytanych raportów i stanów: na starcie sesji (i co 10 min) liczy raporty WATCH (≤ 14 dni, werdykt), USAGE (≤ 14 dni, koszt i udział 2 dni, > 40 %) i pliki stanu innych sesji (≤ 7 dni, tylko nagłówek sekcji „## Stan”) spod ~/.claude/kanon/; jedna linia nad promptem, /skrzynka z przyciskami Wczytaj (prompt.fill) i Przeczytane ($.store). Polecenie „wczytaj <ścieżka>” liczy się jako przeczytane (D6). Zero modelu, zero sieci, zero zapisu plików.

stelmach-systems/claude-mods

Warsztat modów Claude Code: narzędzie modelu `szukaj` (JSDoc i sygnatura zdarzenia, metody `$.…` albo typu z lokalnego claude-code/index.d.ts po nazwie, indeks budowany leniwie, wynik ≤ 2048 znaków — zamiast grep po 16 tys. linii) i komenda /warsztat [ścieżka] (dla katalogu modu kolejno: `claude plugin validate .` i `claude plugin test .` na CLI ~/.local/bin i na binarce desktopu, potem `tsc --noEmit -p .`; tabela binarka × krok z OK/FAIL/czas/liczbą testów, przyciski Powtórz i Kopiuj błędy). Zamknięta lista argv, bez shella, cwd tylko w ~/Developer/claude-mods* albo katalogu sesji. Zero modelu, zero sieci, zero zapisu plików.

geonhos/claude-code-mod

Dancing kaomoji above the prompt, one per harness worker: who is working, who is idle, who needs help.

starecz/karea-agentic

Shows whether your active Karea task is in sync with what this session did: hint-line chip, behind band, changed-in-Karea alerts, task pane, sync receipts, a /ksync command and the Karea skill

ray-manaloto/knowledge-base

Denies a delegated agent lane any write to this repo's Claude settings files.

kcsujeet/kc-claude-kit

Thumbnails of the images you paste, above the prompt: the real picture in Ghostty and kitty, colored cells in other terminals, and the real file behind a copied Finder icon

kcsujeet/kc-claude-kit

A row of pills above the prompt: model, git branch, context fill, 5h and 7d rate limits with reset times, input/output/cache tokens, and session cost, with usage figures turning amber from 70% and red from 90%

Jvrd97/claude-keep-going

Retries a turn that died on a transient API error (500, 529 overloaded, timeouts, connection resets) with backoff, and resumes work when the usage limit resets.

someonewithpc/claude-mod-keep-going

Keeps unattended sessions going: continues after usage limits reset, retries API overload with backoff, resumes cut-off responses, and compacts idle sessions before the prompt cache expires.

ivan-szz/keep-or-undo

Per-session log of Claude's file changes, with Keep and Undo per file.

HitEmUp/keepcachewarm

Zeigt über dem Eingabefeld, wie lange der Prompt-Cache noch warm ist, mit Kontext, Limits und Kosten, und hält ihn auf Wunsch in Pausen warm.

luwux/keepwarm

Idle-session guard for Claude Code's 1-hour prompt cache (macOS): about 50 idle minutes after the last turn of a large session it notifies you, optionally after a cache-read fork judges whether to compact, keep warm or leave it; every request and change it makes is audited. Derived from cache-tax (MIT, Karan Bansal).

SeKondBrainAILabs/kemory-panel

Kemory inside Claude Code: a status line that says whether memory works, the memories in context above the prompt, transcript cards, and a panel to review, rate, save and delete memories. Auto-approves read-only Kemory tools. Requires the kemory plugin.

ajbarea/keryx

Gives Claude Code a voice: after each reply, speaks a one- or two-sentence gist through a local Kokoro voice, shortened by a local Ollama model.

shiran-yu/keyfob

Keep your API keys and tokens out of your Claude Code conversations: tokens pasted as keyfob: lines are stored before the model sees them, Claude asks for a key through the /keyfob panel, a guard stops Bash commands that would print one, and each command gets only the keys it needs from the OS keystore

sattva2020/keystone

Intent-first workflow sized to the change: S (just do it and check), M (short plan with Risks), L (intent → plan → independent check → implement → verify). Built on Anthropic's AI-native SDLC playbook and Claude Code best practices; the plan is the spec.

kikostefanov-lab/claude-code-mods

Kiko boxes every turn above your prompt: Knowledge In, Knowledge Out, K.O.!