Genome Report

Analyze 23andMe raw genome data to generate comprehensive health, trait, and family comparison reports. Supports cardiovascular, cognitive, metabolic, pharma...

Security Scan

The skill's code, data, and runtime instructions match the stated purpose (analyzing 23andMe raw files and producing reports) and request no credentials or installs, but the SKILL.md contains unicode control characters (possible prompt-injection signal) and the source is unknown — review before running on sensitive data.

Purpose & Capability

Name/description align with provided artifacts: a Python script that parses 23andMe v5 raw files and a local SNP database (references/snp_database.json) are present and sufficient for the stated functionality. The skill does not request unrelated binaries, cloud credentials, or external services.

Instruction Scope

SKILL.md instructs running the included Python script and editing the local SNP JSON for updates — this stays within the scope of local genotype analysis. The script reads local genome .txt files, produces text/json/html output, and can scan a directory for family comparisons (glob("*.txt")). It does not appear to make network calls or access external endpoints. Caveat: SKILL.md triggered a unicode-control-chars prompt-injection signal which may be an attempt to manipulate LLM behavior; inspect the full SKILL.md for hidden/control characters before trusting automated agents.

Install Mechanism

No install spec; skill is instruction-only with bundled Python script and data. No downloads, package installs, or archive extraction are declared.

Credentials

The skill requires no environment variables, credentials, or config paths. That is proportionate to its local file-processing purpose.

Persistence & Privilege

Skill does not request always: true and does not modify other skills or system-wide settings. It runs as-invoked and writes report files to local paths (output_path), which is expected behavior.

Scan Findings in Context

[unicode-control-chars] unexpected: The scanner found unicode control characters in SKILL.md. These characters are not necessary for a README/instruction file and can be used to manipulate LLM prompt parsing or hide content. Treat as suspicious: open SKILL.md in a hex-aware editor and remove or inspect any non-printable characters before running in an automated agent.

Assessment

This skill appears internally consistent for local analysis of 23andMe raw files: it parses .txt files, compares genotypes to a bundled SNP database, and writes reports. Before installing or running it, do the following: (1) Inspect the full SKILL.md and the entire genome_report.py for hidden/control characters, network calls (urllib/requests), subprocess exec, or obfuscated code — the pre-scan flagged unicode control chars which can hide prompts. (2) Because genomic data is highly sensitive, run this code locally in a sandbox or isolated VM and do not upload genome files to unknown services. (3) Review references/snp_database.json for accuracy and provenance — the interpretation text can be edited and is not medical advice. (4) If you plan to let an autonomous agent invoke the skill, consider disabling autonomous invocation or restricting the agent's access to genome files; otherwise an agent could process private files without explicit user consent. (5) If you need stronger assurance, ask the publisher for source provenance, a signed release, or run static analysis and a full audit of the script (the provided snippet was truncated in the manifest output; verify the full file).

Analyze 23andMe v5 raw data files and generate health/trait reports with risk scoring.

Usage

python3 skills/genome-report/scripts/genome_report.py <genome_file.txt> [options]

Options

Flag	Values	Default	Description
--category	all|cardio|cognitive|metabolic|pharma|athletic|traits	all	Category filter
--output	text|json|html	text	Output format
--output-path	filepath	auto	HTML output path
--family DIR	directory path	—	Family comparison mode

Examples

# Full report to console
python3 skills/genome-report/scripts/genome_report.py ~/my_genome.txt

# HTML report for one category
python3 skills/genome-report/scripts/genome_report.py ~/my_genome.txt --category cardio --output html

# Family comparison
python3 skills/genome-report/scripts/genome_report.py --family ~/genomes/ --output html

Input Format

23andMe v5 raw data (tab-separated): rsid chromosome position genotype

Lines starting with # are skipped. Genotypes marked -- (no-call) are excluded.

Output

• Text: Console output with risk bars and color-coded SNP details
• JSON: Structured data with scores and per-SNP results
• HTML: Styled report with risk score cards and color-coded tables

Categories & Coverage

~55 curated SNPs across 6 categories:

• Cardiovascular — blood pressure, cholesterol, cardiac rhythm, CAD risk
• Cognitive — memory, dopamine, BDNF, social cognition, brain volume
• Metabolic — diabetes risk, MTHFR, lactose, alcohol, iron metabolism
• Pharmacogenomics — warfarin, clopidogrel, CYP2D6, CYP2C19 drug metabolism
• Athletic — muscle fiber type, endurance, recovery, injury risk
• Traits — eye color, hair color, earwax, bitter taste, asparagus smell

SNP Database

The SNP reference data lives in references/snp_database.json. Edit this file to add/update SNPs without changing the script. Each entry has: rsid, gene, category, trait, risk_allele, and genotype-specific effect descriptions.

Constraints

• Pure Python 3.9+ — no external dependencies
• Not medical advice — educational/informational only