Borrowing it
Nothing to install: this file belongs to karlkfi/claude-bouncer. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/karlkfi/claude-bouncer/main/CLAUDE.mdgit clone --depth 1 https://github.com/karlkfi/claude-bouncerWrote 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/karlkfi/claude-bouncer/claude-md)<a href="https://agentmods.dev/instructions/karlkfi/claude-bouncer/claude-md"><img src="https://agentmods.dev/badge/instructions/karlkfi/claude-bouncer/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/karlkfi/claude-bouncer/claude-md"><img src="https://agentmods.dev/badge/instructions/karlkfi/claude-bouncer/claude-md.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.02032 | $0.02032 |
| Opus 5 | $0.01016 | $0.01016 |
| Sonnet 5 | $0.00406 | $0.00406 |
| Haiku 4.5 | $0.00203 | $0.00203 |
Grade A, and why
claude-bouncer CLAUDE.md 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 6d 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.
How it starts
The opening of the file, as written. The whole thing — 159 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Working in claude-bouncer
Five Claude Code guard plugins in one marketplace. Each plugin under
plugins/<name>/ is self-contained and has its own CLAUDE.md covering its
rules. Read that one for plugin work. This file covers what is true of the repo
as a whole. Releasing is one of those things: it moves the root manifest and
the root README, so the runbook is docs/development/release-process.md and
there is no per-plugin one. The backlog is another — one store for all five,
below.
Never edit a vendored library copy
lib/ holds the shared modules: bouncer_parse.py, the shell parser, and
bouncer_grants.py, the session-grant store. Every plugin carries a copy of
each at plugins/<name>/lib/, written by scripts/sync-lib.py, which syncs
every module named in its MODULES tuple. Adding a module there is what makes
it shared — a file dropped in lib/ and not named is vendored nowhere.
Edit the root copy, then run make sync. A vendored copy is overwritten
without warning, so an edit made there is lost at the next sync, and until then
the guards disagree about what a command means. make sync-check is a CI gate
and runs before the tests: a plugin suite passing against a stale copy is the
failure it exists to catch.
The copies exist because Claude Code copies a plugin into
~/.claude/plugins/cache/<marketplace>/<plugin>/<version>/ at install time and
a path climbing out of the plugin root does not resolve there. Symlinks to a
marketplace sibling are dereferenced for a git-hosted install but skipped for
--plugin-dir and local-path installs, which is how this repo gets tested, so
vendoring is the only form that behaves the same everywhere.
What belongs in the shared parser
A primitive belongs in lib/ when the guards must agree on it. Tokenizing a
command is one. Segmentation is not: foreground-guard needs to know which
segment was backgrounded, exit-status-guard needs the operator that joined two
commands, and prod-guard judges an unterminated heredoc that bash treats as
data. Forcing one answer onto another guard's question is how five copies of
one lexer drifted apart in the first place.
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.
- 6d ago Changed · +11 lines · +186 tokens per session 04118d07f584
- 10d ago First seen · 148 lines · 1,846 tokens per session scan A 6bd834869554
claude-bouncer CLAUDE.md is an instructions file published in the GitHub repository karlkfi/claude-bouncer (3 stars, last pushed today), licensed MIT. It adds 2,032 tokens to every session, about $0.0102 per session on Opus 5. 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.
Other instructions, from other repositories
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
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 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.
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.