Borrowing it
Nothing to install: this file belongs to NamrataAShettar/claude-sethu. 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/NamrataAShettar/claude-sethu/main/CLAUDE.mdgit clone --depth 1 https://github.com/NamrataAShettar/claude-sethuWrote 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/namrataashettar/claude-sethu/claude-md)<a href="https://agentmods.dev/instructions/namrataashettar/claude-sethu/claude-md"><img src="https://agentmods.dev/badge/instructions/namrataashettar/claude-sethu/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/namrataashettar/claude-sethu/claude-md"><img src="https://agentmods.dev/badge/instructions/namrataashettar/claude-sethu/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 | $0.02116 | $0.02116 |
| Opus 5 | $0.01058 | $0.01058 |
| Sonnet 5 | $0.00423 | $0.00423 |
| Haiku 4.5 | $0.00212 | $0.00212 |
Grade A, and why
claude-sethu 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 10d 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 — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.
sethu — working notes for Claude
sethu is a Claude Code plugin: a UserPromptSubmit hook (hooks/sethu_hook.py →
hooks/_engine.py) that runs terminal commands typed as > cmd / >> cmd, plus a
management CLI (sethu --allow …, --mode …, --runner). Pure stdlib, no deps.
Code map
(By module and function/section, not line numbers, so it survives edits. _engine.py
is banner-commented — # ── <section> ── — grep those to jump.)
hooks/sethu_hook.py— theUserPromptSubmitentry. Cheap fast-path gate (_maybe_sethu, decides withjson/osonly, does NOT import_enginefor a normal prompt); routes a sethu prompt toprocess(); runssethu …management commands as a subprocess;_blockemits the zero-tokendecision:block.hooks/_engine.py— the core, in banner sections:- config —
DEFAULTS,load_config(type + value coercion, whitespace-canonicalizes allow),save_config(atomic via_atomic_write: temp+fsync+os.replace, so a crash can't torn/truncate it —set_cwduses it too),_config_lock(fcntl.flock serializing the CLI's read-modify-write so concurrent writers don't lose updates),config_path. - allow matching —
_matches(exact/prefix),_is_chain_unsafe/_has_unquoted_ops/_SUBST_META(the injection-hardened chain guard). - per-session cwd (cwd mode) —
is_cd,resolve_cd,set_cwd/get_cwd(atomic). - command execution — gated decision (
is_gated,GATED(flag-safe tool set, no flag logic),_DANGER,_pipe_segments(quote-aware|split, so a pipe inside a quoted arg likejq '.a | .b'isn't mistaken for a chain),_LAUNCHERS,_why_refused,gated_list_text); interactive/TUI detection (is_interactive,_REPL,_REPL_BATCH_FLAGS,INTERACTIVE,_looks_full_screen);run_capture(the captured runner — streams via Popen, byte-capped atMAX_CAPTURE_BYTES, kills a runaway); truncation + temp hygiene (_truncate,_out_path,_sweep_temp(7-day GC of sethu's/tmpfiles — output logs, launch scripts, sockets, cwd/lock files; throttled to ~1/hr by the mtime of an empty/tmp/sethu-sweptsentinel, openedO_NOFOLLOW, so it doesn't rescan/tmpevery command),_output_path);launch_in_terminal+_launch_command_script. - persistent shell (shell mode) —
shell_run,_sock_path,_spawn_daemon(+_acquire_spawn_lock/_release_spawn_lock, the L7 spawn guard),kill_daemons,_timeout_msg(talks to_shelld.py). - the core: process one submitted prompt —
process()(the one entry that returns{passthrough|block|context}), plus the header/color helpers_header,_reply,_msg,_c,_color_on,_plain_on(plain/spoken mode:sethu:prefix + words, no|^=^|/glyphs — threaded as aplainarg through the render helpers),_style_cli(the hook colors/plain-swaps the management CLI's captured stdout; HELP/help_textembed NO lead so_leadsupplies the colored icon),_ANSI,ICON. - management CLI —
main(wraps the read-modify-write in_config_lock),_apply_cli_mutations(the per-flag config mutations, run under the lock;_announce_launchopens a terminal AFTER the lock),normalize_argv(subcommand→flag aliasing),_Parser(branded errors),help_text,_print_config.
- config —
hooks/_shelld.py— the persistent-shell daemon (PTY + unix-socket server) that backs shell mode; wire protocol version_PROTO. On a command timeout it verifies the shell recovered (_probe/_recover: Ctrl-C → SIGKILL the foreground pgrp → respawn on a wedge, H3) and byte-caps output (MAX_OUTPUT, ST7).hooks/session_start.py— SessionStart hook (first-run welcome).hooks/run.sh— POSIX-sh launcher: execpython3, or degrade gracefully when it's missing (the tolerant no-python3 block shim).hooks/hooks.json— hook registration..claude-plugin/—plugin.json(version lives here),marketplace.json.tests/test_sethu.py— the whole suite + the coverage table andTestCoverageEnforcement(top of file).docs/— design-guidelines, testing, use-cases.
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.
- 10d ago First seen · 116 lines · 2,116 tokens per session scan A 34bb7d9b72f0
claude-sethu CLAUDE.md is an instructions file published in the GitHub repository NamrataAShettar/claude-sethu (1 stars, last pushed 1mo ago), licensed MIT. It adds 2,116 tokens to every session, about $0.0106 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.