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/parthakker/nfl-analytics/claude-mdgit clone --depth 1 https://github.com/parthakker/nfl-analyticsWrote 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/parthakker/nfl-analytics/claude-md)<a href="https://agentmods.dev/instructions/parthakker/nfl-analytics/claude-md"><img src="https://agentmods.dev/badge/instructions/parthakker/nfl-analytics/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 | $0.02205 | $0.02205 |
| Opus 5 | $0.01103 | $0.01103 |
| Sonnet 5 | $0.00441 | $0.00441 |
| Haiku 4.5 | $0.00220 | $0.00220 |
Grade A, and why
nfl-analytics 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 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.
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 — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.
NFL Analytics Warehouse
Repo: https://github.com/parthakker/nfl-analytics (public, MIT). Local git
identity is Parth's personal email (repo-local config). data/* (except the
three hand-curated JSONs), *.duckdb, and logs/ are gitignored — fresh clones
rebuild via nfl refresh --bootstrap. Commit when Parth asks; follow
/release for the ritual.
Personal NFL analyst project. Questions are answered by running SQL against
nfl.duckdb (DuckDB, repo root) — computed answers, not retrieval. Before
non-trivial SQL, load the warehouse-queries skill (table grid, view catalog,
query patterns); per-table dictionaries live in docs/dictionary/*.md.
How to query
python -c "import duckdb; con=duckdb.connect('nfl.duckdb', read_only=True); print(con.execute('''<SQL>''').fetchdf().to_string())"
Always read_only=True for analysis. Never write to any *.duckdb outside
the build scripts (enforced by permissions.deny).
Commands
nfl <cmd>(orpython -m nfl_analytics.cli <cmd>): refresh / rebuild / views / audit / smoke / weather / train / news / kalshi / fixture. Rebuild =nfl rebuildthennfl views(~60s).- Tests:
pytest tests/unit(fast, no DB) ·pytest -m warehouse(real-DB invariants) ·pytest -m api(TestClient contracts) ·cd web/ui && npm run e2e(Playwright vs live server). Prefer running single tests, not the whole suite. CI runs unit + fixture tiers (NFL_TEST_USE_FIXTURE=1). - Chat answer quality has its own suite:
python scripts/run_chat_evals.py(evals/, 28 questions × 4 moments). Costs real API budget and is manual — re-run the affected--momentslice after editing CLAUDE.md,mcp_server.pyorchat.py, and--compareagainst the previous run dir. - UI rebuild after web/ui edits:
cd web/ui && npm run build. - Jarvis (primary UI): "NFL Jarvis" shortcut or
python web/run_web.py(:8000). Legacy Streamlit lives inlegacy/— frozen, don't extend.
Front end
web/ — FastAPI routers (web/api/) + React/Vite/Tailwind SPA (web/ui/).
Nav is 6 sections + More: Today / · Scores /scores · Teams /teams ·
Players /players · Betting /betting · Learn /knowledge; under More sit
Leaders, Coaches, Refs, H2H, Markets, News. Details are singular
(/team/:code, /player/:gsis, /coach/:name, /matchup/:gameId). Old
paths redirect, unknown paths go to /. Ctrl-K is a moded command palette
(jump / > commands / ? analyst) in components/CommandPalette.tsx.
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 · 139 lines · 2,205 tokens per session scan A abf83614244c
nfl-analytics CLAUDE.md is an instructions file published in the GitHub repository parthakker/nfl-analytics (0 stars, last pushed 4d ago), licensed MIT. It adds 2,205 tokens to every session, about $0.0110 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
streamlit copilot-instructions.md
Copilot instructions for streamlit/streamlit, covering streamlit repo overview, tech stack, folder structure, shell & build policy and make commands.
evidence AGENTS.md
AGENTS.md instructions for evidence-dev/evidence, covering agents.md, this repo is a mirror, layout and scripts.
Financial-API AGENTS.md
Instructions for HiThink-Tech/Financial-API, covering agents.md, 必读顺序, monorepo 边界, 选择接入方式 and api key.
chunkhound AGENTS.md
AGENTS.md instructions for chunkhound/chunkhound, covering chunkhound llm context, projectidentity, modificationrules, keycommands and development.
apple-health-mcp AGENTS.md
Instructions for neiltron/apple-health-mcp, covering repository guidance, project, validate changes, code map and invariants.
WrenAI AGENTS.md
AGENTS.md instructions for Canner/WrenAI: Repository-wide instructions live in .claude/CLAUDE.md. Please refer to that file for shared build commands, architecture, and conventions.