token-guard

A Claude Code add-on that monitors how much text the agent reads and how many helper agents it starts. It warns about very large files, limits repeated file reads, estimates token use, and caps helper-agent launches.

In plain words
What is it for?
Use it to control reading-heavy coding tasks, track approximate token usage, and limit how many subagents Claude Code can start in one session.
Why use it?
It helps prevent an automated coding session from using up its token allowance unexpectedly or spending it on repeated, oversized reads.

Plugin for Claude Code

Install

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.

Claude Code
/plugin marketplace add yurukusa/cc-safe-setup
agentmods
npx agentmods add plugins/yurukusa/cc-safe-setup/token-guard
Clone the repo
git clone --depth 1 https://github.com/yurukusa/cc-safe-setup

Made for: Claude Code.

Per session not measured What this adds to a session before it is invoked.
When invoked not measured Not applicable: nothing here is loaded into a session.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Security

Grade A, and why

token-guard 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 2d 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.

plugins/token-guard/.claude-plugin/plugin.json · 54 lines

How it starts

The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.

{
  "name": "token-guard",
  "description": "Token consumption guards for Claude Code. Warns on large file reads (100KB+), limits unique file reads per session, estimates token budget, and caps subagent spawns. From 800+ hours of autonomous operation data.",
  "version": "1.1.0",
  "author": {
    "name": "yurukusa"
  },
  "homepage": "https://yurukusa.github.io/cc-safe-setup/token-book.html",
  "repository": "https://github.com/yurukusa/cc-safe-setup",
  "license": "MIT",
  "hooks": {
    "PreToolUse": [
      {
        "matcher": "Read",
        "hooks": [
          {
            "type": "command",
            "command": "INPUT=$(cat); FP=$(printf %s \"$INPUT\" | { if command -v jq >/dev/null 2>&1; then jq -r '.tool_input.file_path // empty' 2>/dev/null; elif command -v python3 >/dev/null 2>&1; then python3 -c 'import sys,json;d=json.load(sys.stdin);print((d.get(\"tool_input\") or {}).get(\"file_path\") or \"\")' 2>/dev/null; elif command -v node >/dev/null 2>&1; then node -e 'let s=\"\";process.stdin.on(\"data\",d=>s+=d).on(\"end\",()=>{try{process.stdout.write(String((JSON.parse(s).tool_input||{}).file_path||\"\"))}catch(e){}})' 2>/dev/null; else echo __CCHOOK_NO_PARSER__; fi; }); if [ \"$FP\" = \"__CCHOOK_NO_PARSER__\" ]; then W=/tmp/cc-noparser-warned-token-guard-$PPID; [ -f \"$W\" ] || { echo \"WARNING [token-guard]: cannot inspect tool calls - jq, python3 and node are all unavailable, so this plugin is NOT protecting you. Install any one of them.\" >&2; : > \"$W\"; }; exit 0; fi; [ -z \"$FP\" ] || [ ! -f \"$FP\" ] && exit 0; SZ=$(stat -c%s \"$FP\" 2>/dev/null || stat -f%z \"$FP\" 2>/dev/null); [ \"$SZ\" -gt 102400 ] 2>/dev/null && echo \"Warning [token-guard]: $(basename $FP) is $((SZ/1024))KB. Use limit parameter to read only what you need.\" || true"
          }
        ]
      },
      {
        "matcher": "Read",
        "hooks": [
          {
            "type": "command",
            "command": "BUDGET=${CC_READ_BUDGET:-100}; WARN=${CC_READ_WARN:-50}; 

Read the full file on GitHub · 54 lines

Changes

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.

  1. 2d ago First seen · 54 lines scan A 45f7db422f64

Subscribe to this mod's changes

token-guard is a plugin published in the GitHub repository yurukusa/cc-safe-setup (5 stars, last pushed 2d 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 2026-08-31.

Related

Other plugins, from other repositories

ai-engineer

Use this agent when implementing AI/ML features, integrating language models, building recommendation systems, or adding intelligent automation to applications. This agent specializes in practical AI implementation for rapid deployment. Examples:\n\n \nContext: Adding AI features to an app\nuser: "We need AI-po.

ccplugins/awesome-claude-code-plugins · not measured

angelos-symbo

Use this agent when you need to create or convert prompts using the SYMBO (symbolic) notation system. This agent MUST be activated whenever generating SYMBO prompts or converting existing prompts to symbolic format. Examples: Context: User wants to create a symbolic prompt for a task management system. user: '.

ccplugins/awesome-claude-code-plugins · not measured

ai-ethics-governance-specialist

Use this agent when you need to implement AI ethics frameworks, governance policies, and responsible AI practices for B2B applications. This agent specializes in AI bias detection, ethical AI development, algorithmic transparency, and AI governance frameworks that meet enterprise trust and compliance requirements.…

ccplugins/awesome-claude-code-plugins · not measured

agent-manager-skill

Orchestrate multiple local CLI agents via tmux using agent-manager-skill (start/stop/monitor/assign + cron-friendly scheduling).

ccplugins/awesome-claude-code-plugins · not measured

autoresearch

Autonomous skill optimizer using Karpathy's autoresearch methodology. Scores outputs 0-100, mutates prompts, keeps improvements. Includes auto-screenshot and stop-gate hooks.

ccplugins/awesome-claude-code-plugins · not measured

dead-rules-audit

A CLAUDE.md compliance flight-recorder for Claude Code. At SessionStart it parses CLAUDE.md into numbered atomic rules; on every Edit/MultiEdit/Write a PostToolUse hook (async, zero added latency) passively tallies how often each rule was relevant and whether it was followed or violated; at SessionEnd (or on demand…

karanb192/claude-code-hooks · not measured