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.
npx agentmods add instructions/sumonmselim/agentguard/claude-mdgit clone --depth 1 https://github.com/SumonMSelim/agentguardWrote 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/instructions/sumonmselim/agentguard/claude-md)<a href="https://agentmods.dev/instructions/sumonmselim/agentguard/claude-md"><img src="https://agentmods.dev/badge/instructions/sumonmselim/agentguard/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01489 | $0.01489 |
| Opus 5 | $0.00745 | $0.00745 |
| Sonnet 5 | $0.00298 | $0.00298 |
| Haiku 4.5 | $0.00149 | $0.00149 |
Grade F, and why
agentguard CLAUDE.md scanned grade F with 4 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 5d 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
| `block-destructive-ops.sh` | `rm -rf /`, `rm ~`, pipe-to-shell (`curl \| bash`, `wget \| sh`) | Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
When `~/.claude/settings.json` exists, installer merges rather than overwrites: Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
| `block-destructive-ops.sh` | `rm -rf /`, `rm ~`, pipe-to-shell (`curl \| bash`, `wget \| sh`) | Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
| `block-destructive-ops.sh` | `rm -rf /`, `rm ~`, pipe-to-shell (`curl \| bash`, `wget \| sh`) | How it starts
The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Guidance for Claude Code (claude.ai/code) working in this repo.
What this repo is
agentguard installs security guardrails for AI coding agents (Claude Code, Kiro, Cursor, Codex, Grok). Enforces rules at shell hook level — not just instructions. Install target: user home (~/.claude/, ~/.kiro/, ~/.grok/) or project dir (.cursor/), not this repo.
Commands
Preferred: Use the agentguard command (installed by the wrapper or packages).
# Install / manage guardrails
agentguard claude # Claude Code (global)
agentguard kiro # Kiro (global)
agentguard grok # Grok
agentguard cursor # Cursor (project-local, run from CWD)
agentguard all # All agents
agentguard claude --dry-run # Preview without writing
agentguard claude --skills go,aws # With specific skill packs
agentguard cursor --skills go,aws # Cursor full install + skills
agentguard claude --project --skills go # Append skills to CWD only (no hooks)
# Uninstall / check
agentguard uninstall claude
agentguard uninstall claude --dry-run
agentguard check claude
agentguard check all
# Bootstrap (one time only, from a fresh clone — this installs the `agentguard` CLI wrapper)
# ./install.sh claude # after this, use `agentguard claude` etc. for everything
# Tests
bash tests/run_all.sh # All suites
bash tests/claude.sh # Hook logic + Claude install check
bash tests/claude.sh hooks # Hook logic only
bash tests/claude.sh install # Install check only
bash tests/check-sync.sh # Assert instruction files are in sync
Requirements: bash, jq.
Architecture
Hooks (hooks/)
Six shell scripts enforcing rules at tool-call level. Each reads JSON from stdin, exits 2 to block or 0 to allow. Exit codes: 0 = allow, 2 = block (agent sees stderr as feedback), 1 = hook error (also blocks).
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.
- 5d ago First seen · 101 lines · 1,489 tokens per session scan F 9ba6d3a5eef1
agentguard CLAUDE.md is an instructions file published in the GitHub repository SumonMSelim/agentguard (56 stars, last pushed 1mo ago), licensed MIT. It adds 1,489 tokens to every session, about $0.0074 per session on Opus 5. A static security scan graded it F with 4 findings (downloads and executes remote code, reads agent configuration directories, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.