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 skills/danielrosehill/claude-data-analyst-plugin/forensic-sweepnpx skills add danielrosehill/Claude-Data-Analyst-plugin --skill forensic-sweepgit clone --depth 1 https://github.com/danielrosehill/Claude-Data-Analyst-pluginWrote 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/danielrosehill/claude-data-analyst-plugin/forensic-sweep)<a href="https://agentmods.dev/skills/danielrosehill/claude-data-analyst-plugin/forensic-sweep"><img src="https://agentmods.dev/badge/skills/danielrosehill/claude-data-analyst-plugin/forensic-sweep.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 | $0.00067 | $0.01556 |
| Opus 5 | $0.00034 | $0.00778 |
| Sonnet 5 | $0.00013 | $0.00311 |
| Haiku 4.5 | $0.00007 | $0.00156 |
Grade A, and why
forensic-sweep 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 6d 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 — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Forensic Sweep
Real-world data is messy. Clean data is often a clue that something upstream made it that way — and that "something" shapes any conclusion downstream. This skill sweeps for those tells and reports them so the user can decide whether the cleaning is legitimate (a curated public dataset) or misleading (aggregation that hides the interesting variance).
Inputs
- Path to a dataset (CSV / Parquet / Excel / DuckDB table).
- Optional: known provenance (where the data came from, any processing steps the user is aware of). Helps calibrate suspicion.
Recommended CLI tooling
duckdb— distinct counts, null counts, distribution quantiles, digit-frequency tests.uv run --with pandas --with scipy python -c '...'— Benford's law, Shapiro/KS normality, duplicate-run detection.
What to look for
1. Impossible tidiness
- Zero nulls everywhere in a dataset that represents a real-world process that normally has some missingness (surveys, logs, transactions). Real data almost always has some nulls. Total absence suggests imputation or a dropna upstream.
- No duplicates at all in transactional or event data where repeats are natural.
- All strings trimmed, lowercased, single-spaced uniformly — evidence of a normalisation pipeline.
- Dates all in the same exact format, no parsing errors — a cleaner has been through.
2. Imputation fingerprints
- A spike at the median, mean, or mode of a column (frequency of one value disproportionate to the rest of the distribution).
- A spike at zero in a column where zero is a plausible imputation placeholder but not a natural value.
- Values like
-999,-1,9999,N/A,UNKNOWN,MISSINGappearing frequently — sentinel values for missingness. - Categorical columns with an "Other" / "Unspecified" bucket at suspiciously high share (>15%).
3. Smoothing / aggregation
- Numeric distributions that are too normal — run Shapiro-Wilk or Kolmogorov-Smirnov vs. the fitted normal. Real-world measurements are rarely textbook-normal.
- Low kurtosis and no outliers at all (e.g. all values within 2σ of the mean). Real data has tails.
- Time series with implausibly smooth trajectories — check first differences; if they're tightly distributed with no spikes, suspect a moving average or spline fit.
- Rounded numbers in clusters (e.g. everything to the nearest 5 or 10) — rounding or binning upstream.
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.
- 6d ago First seen · 106 lines · 67 tokens per session scan A cfbdba23e4fc
forensic-sweep is a skill published in the GitHub repository danielrosehill/Claude-Data-Analyst-plugin (11 stars, last pushed 4mo ago), licensed MIT. It adds 67 tokens to every session and 1,556 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
sn-search-academic
用于学术调研、论文精读、相关工作梳理、百科知识查询和引用链追溯。.
ppt-analysis
PPT (.pptx/.ppt) 全量解析。覆盖：所有 slide 文本/表格/图表提取、嵌入图片 caption、纯图片 slide 渲染识别、数据标签提取。.
sn-image-base
Base-layer skill for the SenseNova-Skills project, providing low-level APIs for image generation, recognition (VLM), and text optimization (LLM). This skill does not preprocess inputs; it only calls backend services and returns results. This skill is not user-facing and is intended for upper-layer skills only.
category-statistics
提取指定类别列并统计各类别数量与占比，生成高分辨率的柱状图、饼图等组合可视化报告，适用于分类数据的分布情况分析。.
large-file-conditional-formatting
根据Excel总行数自动切换Parquet加速读取，计算特定维度的时间序列平均值，并使用openpyxl输出带有条件格式（如低于均值标绿）和自定义样式的分析报告。.
trend-analysis
基于多维度数据进行分级评估与趋势预测，通过设定差异化增长率计算预测值，并生成对比可视化图表，适用于绩效评估、目标设定等场景。.