What is hvtracker CLAUDE.md?

A set of project instructions for HVTracker, a website that records and compares how much people trust AI agents. It explains the project's rules, generated files, development process, and required checks.

Use it when changing HVTracker code, regenerating website output, or running the project's tests and checks.

It tells coding agents how to work in the repository safely and how to verify changes before opening a pull request.

How YugantM/hvtracker configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to YugantM/hvtracker. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/YugantM/hvtracker/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/YugantM/hvtracker

Made for: Claude Code.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/yugantm/hvtracker/claude-md/github.svg)](https://agentmods.dev/instructions/yugantm/hvtracker/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/yugantm/hvtracker/claude-md"><img src="https://agentmods.dev/badge/instructions/yugantm/hvtracker/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.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/yugantm/hvtracker/claude-md"><img src="https://agentmods.dev/badge/instructions/yugantm/hvtracker/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 2,720 This file is loaded in full into every session.

When invoked 2,720 The same file — it is already loaded in full.

Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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	$0.02720	$0.02720
Opus 5.5	$0.01088	$0.01088
Sonnet 5.5	$0.00544	$0.00544
Haiku 4.5	$0.00272	$0.00272

Measured 25d ago against content hash 2fef4edfb9fa, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

hvtracker 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 25d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- Prod is read-only checkable: `curl -A "Mozilla/5.0" https://hvtracker.net/healthz`

CLAUDE.md · 154 lines

How it starts

The opening of the file, as written. The whole thing — 154 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

HVTracker (hvtracker.net) — AI-agent trust registry (HVTrust scores, grades A–D). FastAPI + static-site generator, deployed on Railway. This file is the session bootstrap: trust it instead of re-discovering the repo; verify only what you change. Milestone history lives in docs/changelog.md — record new milestones there.

Hard rules

One task = one branch = one PR off latest main. main is PR-only, squash-merge (linear history). main == production (since #225).
Merging ≠ deploying. Deploy only when the owner says so, only from main, following the runbook below. Never run mutating railway commands otherwise.
Never hand-edit generated output (agents/, ecosystem/, org/, data/, sitemap.xml, index.html, blog/, compare/*-vs-*/, changes/) — change the generator + re-render.
Never change production rank without an evidence gate (upset review); scoring changes ship as separate visible slices, never silent reweights.
No <title>/meta-description changes except as a deliberate, measured SEO batch (the #114 churn lesson).
Monetization on hold (visa): no billing/paid-tier code.
output/history/*.json daily snapshots are irreplaceable IP — never delete. A snapshot that must not be compared against goes in PARTIAL_SNAPSHOT_DATES (fetch_and_build.py, mirrored in app.py).

Gates — every PR, all green

./scripts/gates.sh

Runs ruff → pytest → --render-only → validate_html.py with the .venv interpreter (bare python isn't on PATH; system python3 lacks the deps), then restores the artifacts the render churns (data/render_state.json, og-v2.png, scorecard-cache.json) unless they were already dirty.
CI runs the same on every PR (the render smoke too, unless MCP-only), plus compileall, pip-audit and shellcheck.
A local render reports a board-churn invariant violation by design; check production's /data/build_report.json, not the local one.
Tests need no Postgres (db.py falls back to agents.json); conftest sets HVT_BOOT_REFRESH=0 so app startup never spawns real refreshes under pytest.
tests/test_predeploy_check.py fails the PR if a runtime module or BASE_DIR asset isn't COPY'd by the Dockerfile (it copies by filename).

Read the full file on GitHub · 154 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.

25d ago Changed · +3 lines · +65 tokens per session 2fef4edfb9fa
26d ago Changed · +2 lines · +33 tokens per session e784bc7eb900
27d ago Changed · +1 lines · +39 tokens per session 0325a09c7ead
1mo ago Changed · +2 tokens per session aad8d596c1a8
1mo ago Changed · +4 lines · +87 tokens per session 6dfa82190711
1mo ago Changed · +11 lines · +215 tokens per session 9a441062c37c
1mo ago Changed · +7 lines · +139 tokens per session 87146e79d212
1mo ago Changed · +6 lines · +115 tokens per session 6f7034904454
1mo ago Changed · +12 lines · +250 tokens per session 65f1b987cb0b
1mo ago First seen · 108 lines · 1,775 tokens per session scan A 0f2056f7ad32

Subscribe to this mod's changes

hvtracker CLAUDE.md is an instructions file published in the GitHub repository YugantM/hvtracker (6 stars, last pushed yesterday), licensed MIT. It adds 2,720 tokens to every session, about $0.0109 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

AI Safety a2a ai-agents leaderboard mcp open-source

hvtracker CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
hvtracker CLAUDE.mdYugantM/hvtracker	Instructions file	6	yesterday	MIT	A	2,720	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,649	GitHub Copilot
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
vscode ai-customization.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	1,907	GitHub Copilot

hvtracker: Instructions file for Claude Code