Best Claude Code Mods (Tested Oct 2026): 20 Worth Installing

#1 Replay Theater – Best for a step-by-step diff viewer for the last turn

Score 94/100 · tested on Claude Code 2.1.310 (2026-10-24) · Intercepts tool calls or prompts

What it does

Replay Theater is a community mod by Hamza Zafar (hamzafer), part of his `claude-code-mods` collection. It works as a small Claude Code diff viewer for the last turn. While Claude edits files, the mod notes each `Edit` and `Write` call: the file, the old text and the new text. When the turn ends, you get a toast, `replay-theater: 2 edits last turn. Run /replay`, and `/replay` opens a pane called Replay Theater.

What we observed

Automated check-compat run on hamzafer/claude-code-mods@77f1ebc/mods/replay-theater. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 3 pass | 0 fail.

Cons

• Edits made through shell commands are not recorded, only Edit and Write calls
• In the onboarding test the p, n and q hotkeys went to the prompt box until the pane had focus

Score breakdown

Install

/plugin install replay-theater --marketplace hamzafer/claude-code-mods

Replay Theater mod page: permissions, setup and FAQ →

#2 Blast Radius – Best for previewing rm -rf and risky commands before they run

Score 86/100 · tested on Claude Code 2.1.310 (2026-10-24) · Runs shell commands

What it does

Blast Radius is a community mod by Hamza Zafar (hamzafer), from his `claude-code-mods` collection. When Claude is about to run a risky shell command through the `Bash` tool, the mod holds it and shows you what it would change, like a Claude Code rm -rf preview, then waits for you to press Cancel or Proceed.

What we observed

Automated check-compat run on hamzafer/claude-code-mods@a6b972c/mods/blast-radius. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 17 pass | 0 fail.

Cons

• Only holds three kinds of command (recursive rm, force push, migrations); a plain rm without -r or rm hidden behind a variable is not caught
• In claude -p and other headless sessions held commands are refused outright

Score breakdown

Install

/plugin install blast-radius --marketplace hamzafer/claude-code-mods

▶ Play demo

Blast Radius mod page: permissions, setup and FAQ →

#3 Flightdeck – Best for an agent dashboard for the terminal

Score 74.5/100 · tested on Claude Code 2.1.310 (2026-10-24) · Intercepts tool calls or prompts

What it does

Flightdeck is a community mod by Stephen Casella (scasella) that works as a Claude Code agent dashboard: one pane that shows what your session and its agents are doing right now. It gathers the model name and effort, a context gauge, session cost, rate-limit windows when your plan reports them, every permission check, subagent cards, and a receipt for the last turn.

What we observed

Automated check-compat run on scasella/claude-flightdeck@6150b9f. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 28 pass | 0 fail.

Cons

• The pane does not open by itself below 144 columns
• Connectors and clocks are static on VS Code and mobile

Score breakdown

Install

/plugin install flightdeck --marketplace scasella/claude-flightdeck

▶ Play demo

Flightdeck mod page: permissions, setup and FAQ →

#4 Next Steps – Best for prompt suggestions after every turn

Score 71.25/100 · tested on Claude Code 2.1.310 (2026-10-24) · Network or model calls

What it does

Next Steps is a community mod by Hamza Zafar (hamzafer), part of his `claude-code-mods` collection. It adds Claude Code prompt suggestions: when a turn ends, it shows two or three short prompts you are likely to type next, in a band above the input. For example, after Claude explains a file it might offer `Write a test for the login function`.

What we observed

Automated check-compat run on hamzafer/claude-code-mods@5657c52/mods/next-steps. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 16 pass | 0 fail.

Cons

• Calls Haiku after every completed main-agent turn, which counts toward your usage
• Digit shortcuts only work in an empty prompt box

Score breakdown

Install

/plugin install next-steps --marketplace hamzafer/claude-code-mods

▶ Play demo

Next Steps mod page: permissions, setup and FAQ →

#5 Secret Redactor – Best for redacting keys and emails in prompts and tool output

Score 68.25/100 · tested on Claude Code 2.1.310 (2026-10-24) · Intercepts tool calls or prompts

What it does

Secret Redactor is a community mod by Ray Amjad (ray-amjad), one of the plugins in his `awesome-claude-code-function-hooks` collection. It does Claude Code secret redaction on three paths: what you type, what a tool reads back, and the context blocks attached to the first message. Each secret, email address or public IP address is swapped for a placeholder such as `[REDACTED-SECRET-164d0c98]` before the model sees it.

What we observed

Automated check-compat run on ray-amjad/awesome-claude-code-function-hooks@12b5fea/plugins/secret-redactor. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. No *.test.ts files, so claude plugin test was not run.

Cons

• Restores real values into any tool input, so a placeholder in a curl command sends the real key
• Heuristic detection can miss secrets, and the vault does not survive a restart

Score breakdown

Install

/plugin install secret-redactor --marketplace ray-amjad/awesome-claude-code-function-hooks

▶ Play demo

Secret Redactor mod page: permissions, setup and FAQ →

#6 Context Bar – Best for seeing what fills your context window

Score 68/100 · tested on Claude Code 2.1.310 (2026-10-24) · UI only

What it does

Context Bar is a community mod by Hamza Zafar (hamzafer), part of his `claude-code-mods` collection. It is a Claude Code context bar that answers one question: what is filling my context window? Above the Claude Code prompt it draws a small bordered band with a single stacked bar the width of the terminal. The bar is split into the same categories `/context` uses, such as system prompt, tools, MCP tools, memory files, skills and messages, followed by free space and the auto-compact buffer.

What we observed

Automated check-compat run on hamzafer/claude-code-mods@a6b972c/mods/context-bar. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 6 pass | 0 fail.

Cons

• The per-category breakdown is only available after Claude Code knows the window size, so the band can be missing until the first prompt

Score breakdown

Install

/plugin install context-bar --marketplace hamzafer/claude-code-mods

▶ Play demo

Context Bar mod page: permissions, setup and FAQ →

#7 Sub Meter – Best for subscription usage in the footer

Score 68/100 · tested on Claude Code 2.1.310 (2026-10-24) · UI only

What it does

Sub Meter is a community mod by Miguel Ángel Durán (midudev), one of the plugins in his better-claude collection. It shows your Claude Code subscription usage at the bottom right of the prompt footer, on the same row as the permission mode. For each window it draws a bar, the percentage used, and the time until that window resets:

What we observed

Automated check-compat run on midudev/better-claude@346837e/plugins/sub-meter. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 26 pass | 0 fail.

Cons

• Needs a Claude subscription that reports rate limits; shows nothing on API-key billing
• Redraws every 8 seconds, so the countdown can look stale

Score breakdown

Install

/plugin install sub-meter --marketplace midudev/better-claude

Sub Meter mod page: permissions, setup and FAQ →

#8 Usage Meter – Best for 5-hour and 7-day plan limit bars

Score 68/100 · tested on Claude Code 2.1.310 (2026-10-24) · UI only

What it does

Usage Meter is a community mod by Hamza Zafar (hamzafer), from his `claude-code-mods` collection. It draws one line above the prompt with your Claude Code plan limits (the rate-limit windows) and the session's cost, for example `⏱ 5h █▆▁▁▁ 34% · resets 2h10 · 7d ███▁▁ 61% · $0.42 session`.

What we observed

Automated check-compat run on hamzafer/claude-code-mods@a6b972c/mods/usage-meter. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 11 pass | 0 fail.

Cons

• Plan limit bars need a Claude subscription; with an API key only cost is shown

Score breakdown

Install

/plugin install usage-meter --marketplace hamzafer/claude-code-mods

▶ Play demo

Usage Meter mod page: permissions, setup and FAQ →

#9 Prismantis Colorful Replies – Best for colorful reply themes

Score 67.5/100 · tested on Claude Code 2.1.310 (2026-10-24) · Intercepts tool calls or prompts

What it does

Prismantis is a community mod by Nahum Litvin (NahumLitvin) that redraws Claude's replies in the transcript with color and structure, in one of 15 Claude Code themes. Markdown tables get colored headers, boxed cells and a double rule under the header, and they shrink to fit the terminal. Code blocks get a language header, a copy button and Prism highlighting for 24 languages, and shell blocks color the command word, flags, strings and comments. Code blocks tagged `mermaid` are drawn as colored text art: flowcharts, sequence, state, class and ER diagrams, plus bar and line charts from `xychart-beta`.

What we observed

Automated check-compat run on NahumLitvin/prismantis@3192fc7. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 138 pass | 0 fail.

Cons

• Attaches a ~190-token note to every prompt you type unless diagramHints is turned off
• Ships whitespace-minified vendored bundles of Prism and a Mermaid renderer

Score breakdown

Install

/plugin install prismantis --marketplace NahumLitvin/prismantis

▶ Play demo

Prismantis Colorful Replies mod page: permissions, setup and FAQ →

#10 Collision Guard – Best for running several Claude Code sessions on one repo

Score 66.5/100 · tested on Claude Code 2.1.310 (2026-10-24) · Runs shell commands

What it does

The Claude Code collision guard mod, Collision Guard, is a community mod by Nate Herk (nateherkai), part of his `claude-code-mods` collection. It is for people who run Claude Code parallel sessions in one repository, where two chats can end up changing the same file without knowing about each other. Before Claude edits or writes a file, the mod checks whether another open chat changed that same file in the last 30 minutes.

What we observed

Recorded in the demo project with a throwaway ledger standing in for a second chat ("Refactor the login flow") that edited src/index.js 4 minutes earlier. When Claude started Update(src/index.js), the Collision question appeared with Proceed, Move to a worktree, Cancel and a free-text answer; choosing Cancel refused the edit and Claude reported why and asked how to continue. claude plugin validate: no errors; the author ships no tests, so claude plugin test had nothing to run.

• Asks instead of hard-blocking, with a git worktree option for tracked files
• No network, no model calls; one local git ls-files check

Cons

• Only sees other chats that also have the mod loaded
• Stores each chat's first prompt (up to 80 characters) in ~/.claude/mods-data/, readable by other local processes
• No author tests; behavior was checked only through the recording

Score breakdown

Install

/plugin install collision-guard --marketplace nateherkai/claude-code-mods

▶ Play demo

Collision Guard mod page: permissions, setup and FAQ →

#11 Token Weather – Best for a context window forecast in the status line

Score 66.5/100 · tested on Claude Code 2.1.310 (2026-10-24) · Runs shell commands

What it does

Token Weather is a community mod by Hamza Zafar (hamzafer), part of his `claude-code-mods` collection. It gives you a Claude Code context window forecast on one line above the prompt: a weather word and icon for how full the window is, the percentage, the token count against the window size, a sparkline of the last turns, and a countdown for how long the prompt cache stays warm.

What we observed

Automated check-compat run on hamzafer/claude-code-mods@5c925b0/mods/token-weather. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 24 pass | 0 fail.

Cons

• Cache countdown detection can fail for very long or moved project paths; set cacheTtl by hand then
• Narrow terminals under 60 columns show only part of the forecast

Score breakdown

Install

/plugin install token-weather --marketplace hamzafer/claude-code-mods

▶ Play demo

Token Weather mod page: permissions, setup and FAQ →

#12 Reflect – Best for saving your corrections to CLAUDE.md

Score 64.25/100 · tested on Claude Code 2.1.310 (2026-10-24) · Network or model calls

What it does

The Claude Code reflect mod, Reflect, is a community mod by Bayram Annakov (BayramAnnakov), the `mod` plugin of his `claude-reflect` repository. It turns the corrections you make while working into CLAUDE.md rules: when you type "no, use pnpm" or "always answer in English", the mod notices the reusable part and offers it in a band above the prompt, so you can turn it into one of your CLAUDE.md rules at the moment you said it.

What we observed

Recorded in the demo project: the prompt "remember: use pnpm, not npm, in this repo" produced the band reflect · "use pnpm, not npm, in this repo" with the buttons Save to ~/.claude/CLAUDE.md, This repo instead, Edit and Skip above the prompt. The recording cannot press buttons, so writing to CLAUDE.md was not exercised on screen, and the "remember:" prefix skips the model check, so the Haiku classifier path was not shown either; both are covered by the author's tests (claude plugin test: 10 pass, 0 fail). claude plugin validate: no errors or warnings for the plugin.

• Nothing is written to CLAUDE.md until you press a Save button, inside a marked section
• Rules are rewritten as one short imperative line and repeats are counted instead of duplicated
• Author ships tests (10 pass on 2.1.310)

Cons

• Each eligible short prompt costs one small model call (Haiku by default) that counts toward your usage
• The classifier and the save path were checked by the author's tests, not seen in the recording
• Lives in the repo's mod/ folder; the repo root is a different, older Python plugin, which is easy to install by mistake

Score breakdown

Install

/plugin install reflect-mod --marketplace BayramAnnakov/claude-reflect

▶ Play demo

Reflect mod page: permissions, setup and FAQ →

#13 Agent Radar – Best for a subagent monitor above the prompt

Score 62/100 · tested on Claude Code 2.1.310 (2026-10-24) · Intercepts tool calls or prompts

What it does

Agent Radar is a community mod by Hamza Zafar (hamzafer), part of his `claude-code-mods` collection. It is a Claude Code subagent monitor: when Claude hands work to a subagent, the band above the prompt gets one live line for it, so you can see what each one is doing without opening anything.

What we observed

Automated check-compat run on hamzafer/claude-code-mods@5657c52/mods/agent-radar. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 2 pass | 0 fail.

Cons

• Only lists subagents it saw spawn in the current session; earlier ones need /radar to pull in

Score breakdown

Install

/plugin install agent-radar --marketplace hamzafer/claude-code-mods

▶ Play demo

Agent Radar mod page: permissions, setup and FAQ →

#14 Block Dangerous Commands – Best for a one-line hard block on rm -rf / and DROP TABLE

Score 62/100 · tested on Claude Code 2.1.310 (2026-10-24) · Intercepts tool calls or prompts

What it does

This mod lets you block dangerous commands Claude Code is about to run through its `Bash` tool. Before any shell command executes, it is checked against a short list of rules. A match is refused, Claude is told which rule fired and why, and you see a toast with the rule name. Safe commands pass through untouched.

What we observed

Automated check-compat run on mods/cm-block-dangerous-commands. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 5 pass | 0 fail.

Cons

• Pattern-based: only the Bash tool is checked, and a command written to avoid the patterns gets through
• Edit and Write calls are not covered

Score breakdown

Install

/plugin install cm-block-dangerous-commands --marketplace rotbit/claudemods-marketplace

▶ Play demo

Block Dangerous Commands mod page: permissions, setup and FAQ →

#15 Loop Detective – Best for catching repeated tool failures early

Score 62/100 · tested on Claude Code 2.1.310 (2026-10-24) · Intercepts tool calls or prompts

What it does

Loop Detective is a community mod by YeonwooSung (YeonwooSung), part of the `my-claude-code-mods` collection. It watches for Claude Code repeated failures: the same Bash command, run again and again, failing the same way. After three identical failures it opens a case and tells you in three places: the status line reads `loop: 1 active · max 3 failures`, a toast names the case, and `/loop-report` prints the evidence.

What we observed

Automated check-compat run on YeonwooSung/my-claude-code-mods@82f9bf6/loop-detective. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 41 pass | 0 fail.

Cons

• Only byte-identical repeated commands form a streak; changing output such as timestamps breaks detection
• Its status text replaces the ? for shortcuts hint in the footer

Score breakdown

Install

/plugin install loop-detective --marketplace YeonwooSung/my-claude-code-mods

Loop Detective mod page: permissions, setup and FAQ →

#16 Context Meter – Best for a minimal context usage status line

Score 61/100 · tested on Claude Code 2.1.310 (2026-10-24) · UI only

What it does

Context Meter is a Claude Code context usage status line. It writes one compact line into the status line: `ctx 42% · 5h 23% · 7d 61.5% · $1.20`. That is how full the context window is, how much of your 5-hour and 7-day plan windows you have used (on a subscription), and what the session has cost so far. When the context reaches 80 percent, a toast warns you once so you can `/compact` or `/clear` before Claude starts losing earlier detail.

What we observed

Automated check-compat run on mods/cm-context-meter. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 4 pass | 0 fail.

Cons

• Shares the status line with other status mods or a custom statusLine command and can be crowded out
• No plan-limit figures when using an API key

Score breakdown

Install

/plugin install cm-context-meter --marketplace rotbit/claudemods-marketplace

▶ Play demo

Context Meter mod page: permissions, setup and FAQ →

#17 Model Effort Band – Best for seeing the effort level above the prompt

Score 61/100 · tested on Claude Code 2.1.310 (2026-10-24) · UI only

What it does

Model Effort Band is a community mod by YeonwooSung (YeonwooSung), part of his `my-claude-code-mods` collection. It puts one band directly above the Claude Code prompt that answers two questions at a glance: which model is selected, and which Claude Code effort level the last main request used.

What we observed

Automated check-compat run on YeonwooSung/my-claude-code-mods@82f9bf6/model-effort-band. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 59 pass | 0 fail.

Cons

• Shows the effort at its own position in the hook chain; a later mod can still change the request
• Effort reads unknown after /model, /effort or any config change until the next prompt

Score breakdown

Install

/plugin install model-effort-band --marketplace YeonwooSung/my-claude-code-mods

Model Effort Band mod page: permissions, setup and FAQ →

#18 Pin Memo – Best for pinned notes and todos by the prompt

Score 61/100 · tested on Claude Code 2.1.310 (2026-10-24) · UI only

What it does

Pin Memo is a community mod by YeonwooSung (GitHub), one plugin in the `my-claude-code-mods` collection. It gives Claude Code pinned notes: you type `/pin Remember to run tests` and the note stays in a teal bordered band above the prompt, so it is still in view after the conversation scrolls past it.

What we observed

Automated check-compat run on YeonwooSung/my-claude-code-mods@82f9bf6/pin-memo. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 29 pass | 0 fail.

Cons

• Notes are kept per folder and session, so a new session or /clear starts an empty list
• No live sync between two processes on the same session; last write wins

Score breakdown

Install

/plugin install pin-memo --marketplace YeonwooSung/my-claude-code-mods

▶ Play demo

Pin Memo mod page: permissions, setup and FAQ →

#19 Token Meter – Best for token usage per turn

Score 61/100 · tested on Claude Code 2.1.310 (2026-10-24) · UI only

What it does

Token Meter is a community mod by Miguel Ángel Durán (midudev), one of the plugins in his better-claude collection. It puts Claude Code token usage where you can see it without running a command: after each turn, the hint line under the prompt ends with that turn's tokens and the session's running total.

What we observed

Automated check-compat run on midudev/better-claude@346837e/plugins/token-meter. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 4 pass | 0 fail.

Cons

• Draws nothing until the first turn reports usage
• Another mod that replaces the hint line can hide it

Score breakdown

Install

/plugin install token-meter --marketplace midudev/better-claude

Token Meter mod page: permissions, setup and FAQ →

#20 Context Inspector – Best for finding what wastes your context

Score 60/100 · tested on Claude Code 2.1.310 (2026-10-24) · Writes files

What it does

Context Inspector is a community mod by YeonwooSung (my-claude-code-mods) that looks for Claude Code context waste. It keeps a ledger of everything that enters the conversation (prompts, replies, tool results, attachments) and answers three questions: what is filling the window now, why the same information keeps coming back, and what a compaction kept or lost.

What we observed

Automated check-compat run on YeonwooSung/my-claude-code-mods@82f9bf6/context-inspector. claude plugin validate (exit 0): ⚠ Found 1 plugin directory warning: | ✔ Validation passed. claude plugin test (exit 0): 102 pass | 0 fail.

Cons

• The ledger starts over after a reload, resume or /clear
• Compaction survival is judged by substring match, so a paraphrased fact counts as lost

Score breakdown

Install

/plugin install context-inspector --marketplace YeonwooSung/my-claude-code-mods

▶ Play demo

Context Inspector mod page: permissions, setup and FAQ →