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 skills add cfagan17/muffed-mcp --skill muffed-figuresgit clone --depth 1 https://github.com/cfagan17/muffed-mcpWrote 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/skills/cfagan17/muffed-mcp/muffed-figures)<a href="https://agentmods.dev/skills/cfagan17/muffed-mcp/muffed-figures"><img src="https://agentmods.dev/badge/skills/cfagan17/muffed-mcp/muffed-figures.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 5.1 | $0.00112 | $0.01945 |
| Opus 5 | $0.00056 | $0.00972 |
| Sonnet 5 | $0.00022 | $0.00389 |
| Haiku 4.5 | $0.00011 | $0.00194 |
Grade A, and why
muffed-figures 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 7d 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 — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Working with Muffed's verified figures
Muffed serves NFL figures that were verified against a gated panel before they were published. The value is not that the numbers exist — it is that each one arrives with the rank its source gave it, the population it was ranked against, its own attribution, and an as-of date. Handling a response carelessly throws away the only thing that makes it better than a guess.
These are the house rules. They exist because each one has a specific way of going wrong.
1 · Quote verified_sentence
Every successful response from a verified tool carries verified_sentence: a
complete, already attributed claim built from the returned row.
(run_stat_query is the exception — see rule 2.)
"No one posted more rush yards over expected than James Cook with 358.16 in 2025, 1st of 49 (Data via nflverse · NFL Next Gen Stats via nflverse)."
Quote it. Do not rebuild the same claim from the individual fields — that is where a rank gets attached to the wrong figure, or a season to the wrong number. The sentence has already related them correctly.
The prose block at content[0] is written to be read aloud and is safe to use
directly. values{} carries every figure in it, exact and unrounded, when you
need more precision than the prose shows.
2 · Computed figures are a different class from verified ones
Two classes of figure come off this server and they are not interchangeable.
Most tools return figures that were verified before they were published: those
carry verified_sentence and a citation that says verified. run_stat_query
computes a figure on request instead, for questions the panel holds no
pre-computed row for — qualified superlatives, multi-season spans, derived
rates.
Its responses carry computed_sentence rather than verified_sentence, a
method block naming the population, the measure and the query shape, and the
computed_not_verified disclosure, which fires on every one of them.
Quote computed_sentence for those, exactly as rule 1 says to quote
verified_sentence. Where it could matter to a reader, say the figure was
computed on request rather than drawn from the verified panel.
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.
- 7d ago First seen · 166 lines · 112 tokens per session scan A dcdaeddf4914
muffed-figures is a skill published in the GitHub repository cfagan17/muffed-mcp (0 stars, last pushed 20d ago), licensed MIT. It adds 112 tokens to every session and 1,945 once invoked, about $0.0006 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 skills, from other repositories
dcf-model
Build discounted cash flow valuation workbooks in Excel.
grounded-citations
Ground answers and documents in cited, verifiable sources.
hermes-agent-skill-authoring
Author in-repo SKILL.md files: frontmatter and structure.
openhands
Delegate coding to OpenHands CLI (model-agnostic, LiteLLM).
baoyu-article-illustrator
Article illustrations: type × style × palette consistency.
flash-attention
Speed up long-sequence transformer training and inference.