Borrowing it
Nothing to install: this file belongs to JSungMin/vs-token-safer. 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/JSungMin/vs-token-safer/main/CLAUDE.mdgit clone --depth 1 https://github.com/JSungMin/vs-token-saferWrote 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/jsungmin/vs-token-safer/claude-md)<a href="https://agentmods.dev/instructions/jsungmin/vs-token-safer/claude-md"><img src="https://agentmods.dev/badge/instructions/jsungmin/vs-token-safer/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/jsungmin/vs-token-safer/claude-md"><img src="https://agentmods.dev/badge/instructions/jsungmin/vs-token-safer/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 | $0.01879 | $0.01879 |
| Opus 5.5 | $0.00752 | $0.00752 |
| Sonnet 5.5 | $0.00376 | $0.00376 |
| Haiku 4.5 | $0.00188 | $0.00188 |
Grade A, and why
vs-token-safer CLAUDE.md scanned grade A with 1 finding 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 29d 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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- **No console windows on Windows:** every `spawn`/`execFile*`/`execSync` in `server/` passes How it starts
The opening of the file, as written. The whole thing — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
vs-token-safer — Claude rules
Routes an agent's code search through an official language server's index (clangd, a Roslyn LSP,
tsserver, pyright), falls back to tree-sitter / a mined concept dictionary / document sections, and
token-caps every answer to a compact file:line list. Local-only. Ships as an MCP server (vs-search),
a CLI (vts, on the plugin PATH via bin/), and PreToolUse hooks. Config ~/.vs-token-safer, env VTS_*.
This file is loaded into EVERY session and subagent turn here, so it stays short. The per-module design,
the measurements behind each rule, and the rejected alternatives live in docs/ARCHITECTURE.md — read
the section for the code you are changing (Layout / Conventions / Backends / Identity) before editing it.
First, orient (every session)
node eval/run.mjsmust printEVAL PASSEDbefore you change anything.- Context: this file ·
docs/ARCHITECTURE.md· the wiki (wiki_query "vs-token-safer",.omc/wiki/; the Status and TODO page is the live checklist) · memory anchorproject-vs-token-safer.
Identity and the roadmap rule
The layer an agent talks to instead of reading the repository: it answers at the highest precision it can
reach and labels the rung (completenessCert): exact (language server) → syntactic (tree-sitter) →
fuzzy (concept dictionary, no embeddings) → section (docs/config by heading/selector), plus
PARTIAL/INCONCLUSIVE coverage. A feature earns its place only by adding a rung or covering more of the
repo, and must keep the three disciplines: capped output, honest precision label, nothing leaves
the machine.
Map (details: docs/ARCHITECTURE.md → Layout)
server/core.js—runTool()dispatch for every tool, token-cap formatters, savings ledger, discover.server/lsp.js— LSP client.server/backends/index.js— backend spawn configs,pickBackend, CDB dirs.server/treesitter.js+symindex.js— syntactic tier + committable.vts-index.server/concept.js— fuzzy tier.server/textstruct.js— section tier.server/scope.js— indexing scope / preindex.server/policy.js— routing digest + suppression.server/squeeze.js— passthrough compaction.server/compact.js— git/p4 output compaction.server/psearch.js— PowerShell search classifier.server/viz.js/serve.js/dashboard.html— local dashboard (vts serve, 127.0.0.1).hooks/block-code-grep.js— the enforcement hook (Bash/Grep/Glob/Read/Edit/PowerShell).hooks/orchestrator-redirect.js— qvts delegation.hooks/edit-report.js— SessionStart digest.agents/code-locator.md— locate-only subagent (tools restricted).eval/run.mjs— the eval.gamedev-log-analyzer/— static mirror; the source is../rider-mcp-enforcer.
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.
- 29d ago Changed · -592 lines · -17,669 tokens per session 4faa9b03f476
- 1mo ago Changed · +15 lines · +450 tokens per session 29202604c85e
- 1mo ago Changed · +52 lines · +1,466 tokens per session bb5f5f400135
- 1mo ago First seen · 610 lines · 17,632 tokens per session scan A 43c13f80a10a
vs-token-safer CLAUDE.md is an instructions file published in the GitHub repository JSungMin/vs-token-safer (13 stars, last pushed 20d ago), licensed MIT. It adds 1,879 tokens to every session, about $0.0075 per session on Opus 5.5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
repobrain copilot-instructions.md
Copilot instructions for study8677/repobrain, covering github copilot bootstrap instructions and hard rule — query the repobrain hub first.
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).
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.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
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.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).