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/yugantm/hvtracker/claude-mdgit clone --depth 1 https://github.com/YugantM/hvtrackerWrote 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/yugantm/hvtracker/claude-md)<a href="https://agentmods.dev/instructions/yugantm/hvtracker/claude-md"><img src="https://agentmods.dev/badge/instructions/yugantm/hvtracker/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 | $0.11830 | $0.11830 |
| Opus 5 | $0.05915 | $0.05915 |
| Sonnet 5 | $0.02366 | $0.02366 |
| Haiku 4.5 | $0.01183 | $0.01183 |
Grade C, and why
hvtracker CLAUDE.md scanned grade C with 2 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 4d 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.
Reaches for credential fileshighPrivilege escalation
SSH keys, cloud credentials, git-credentials, .npmrc, /etc/shadow: reading these is how a config file becomes a credential leak.
and suggested overwriting ~/.ssh/id_ed25519 (existing Apr-12 key, 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` How it starts
The opening of the file, as written. The whole thing — 607 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.
Hard rules
- One task = one branch (
feat/<id>) = one PR off latestmain.mainis PR-only, squash-merge (linear history), 0 approvals required, CI not a gate. - Merging ≠ deploying. Deploy is manual
railway upfrom a clean worktree. NEVER deploy or run railway commands unless explicitly told. - 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.
- Monetization on hold (visa): no billing/paid-tier code.
output/history/*.jsondaily snapshots are irreplaceable IP — never delete.
Gates — every PR, all three green
python -m pytest && python fetch_and_build.py --render-only && python tests/validate_html.py
- CI additionally runs
ruff check .(Python only), compileall, shellcheck. --render-onlychurns generated artifacts; restore before committing:git checkout -- data/render_state.json og-v2.pngindex.html,methodology.html,output/are gitignored (server-rendered).- Tests need no Postgres:
db.pyfalls back toagents.jsonwhen DATABASE_URL is unset;tests/test_api.pybuilds the site into a tmp OUTPUT_DIR.
Map — grep, don't read wholesale
fetch_and_build.py(~280KB, ~6k lines) — the generator. NEVER read whole; grepdef <name>. Key:compute_weekly_changes(/changes/ + RSS viabuild_changes_rss),compute_movers(daily gainers/losers ticker,limitparam),derive_agent_events(threshold events →recent_events→ bell notifications),_load_prior_snapshot(rank Δ vs yesterday's history snapshot),compute_trust_score_v2(runtime calibration — IS the production trust_score/rank/grade since methodology v4.0).app.py— FastAPI serving, /healthz, 30-min signals-refresh scheduler.auth.py— OAuth/password accounts, watchlist,/api/notifications(derive-on-read fromrecent_eventsindata/latest.json; needs DB for sign-in).auth.js— header widget incl. notification bell.db.py— Postgres layer + file fallback.schema.sqlidempotent DDL.template.html— homepage (leaderboard, Δ column, ticker).templates/*.j2— all other pages. Grade-B color #2c5282 is a design invariant.- Blog post = 4 surfaces:
blog_static/<slug>/, blog_index card,sitemap_urls,blog_feed_items(all wired infetch_and_build.py). - Scorecard data comes from our own CLI scan via the
databranch (not deps.dev).
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.
- 4d ago First seen · 607 lines · 11,830 tokens per session scan C 740c01712e3c
hvtracker CLAUDE.md is an instructions file published in the GitHub repository YugantM/hvtracker (5 stars, last pushed 4d ago), licensed MIT. It adds 11,830 tokens to every session, about $0.0592 per session on Opus 5. A static security scan graded it C with 2 findings (reaches for credential files, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
floe-guard AGENTS.md
AGENTS.md instructions for Floe-Labs/floe-guard, covering agents.md — floe-guard, what floe-guard is, add it to an agent — unattended steps, rules — get these right and the full govern-your-spend workflow (skill).
AUTHENSOR AGENTS.md
Instructions for AUTHENSOR/AUTHENSOR, covering agents.md — authensor integration guide for ai agents, what authensor does for you, quick integration, if you have access to http and if you have access to npm / npx.
ardur AGENTS.md
AGENTS.md instructions for ArdurAI/ardur, covering ardur agent instructions, 0. first action in every new session, 1. what ardur is, and what it does not claim, 2. tl;dr commands and 3. architecture and trust boundaries.
AUTHENSOR CLAUDE.md
Instructions for AUTHENSOR/AUTHENSOR, covering authensor — claude code project context, what is this project, repository structure, key technical conventions and development commands.
ardur CLAUDE.md
Claude Code instructions for ArdurAI/ardur: This repository keeps a single, canonical set of agent instructions in AGENTS.md. This file exists only so that Claude Code loads them automatically; it deliberately holds no rules of its own, so that the two can never drift apart.
ardur GEMINI.md
Gemini CLI instructions for ArdurAI/ardur: This repository keeps a single, canonical set of agent instructions in AGENTS.md. This file exists only so that the Gemini CLI loads them automatically; it deliberately holds no rules of its own, so that the two can never drift apart.