scrutiny-viz

scrutiny-viz is the comparison and reporting component of the broader Scrutiny toolset.

It compares measurement profiles using YAML schemas and can produce:

• normalized JSON from supported raw inputs
• a machine-readable verification JSON
• an HTML report for human inspection
• batch comparison outputs for one-to-many workflows

Scrutiny-viz can be used as a standalone tool or as a later-stage component in a larger Scrutiny workflow.

Main Scrutiny project: https://github.com/crocs-muni/scrutiny

Main workflows

The unified CLI entry point is:

python scrutinize.py -h

Available commands:

• map — convert supported raw inputs into normalized JSON
• verify — compare two normalized JSON files using a schema
• report — render an HTML report from verification JSON
• full — map if needed, then verify, then report
• batch-verify — compare one reference against many profiles

Quick start

Requirements

• Python 3.13+
• pip

Setup

git clone https://github.com/crocs-muni/scrutiny-viz.git
cd scrutiny-viz
python -m venv .venv
source .venv/bin/activate  # Linux / macOS
# or .\.venv\Scripts\Activate.ps1 on Windows PowerShell
pip install -r requirements.txt

Minimal JSON workflow

python scrutinize.py verify \
  -s scrutiny/schemas/TPMAlgTest.yml \
  -r examples/TPM/tpm_example1.json \
  -p examples/TPM/tpm_example2.json \
  -o results/tpm_verify.json

python scrutinize.py report \
  -p results/tpm_verify.json \
  -o tpm_comparison.html

One-command workflow for CSV/JSON inputs

python scrutinize.py full \
  -s scrutiny/schemas/TPMAlgTest.yml \
  -r tests/mapper/test-data/TPMAlgTest/INTC_Intel_10.0.36.1030.csv \
  -p tests/mapper/test-data/TPMAlgTest/INTC_Intel_10.0.36.1030.csv \
  -t tpm \
  --verify-output results/tpm_verify.json \
  --report-output tpm_comparison.html

Batch verification example

python scrutinize.py batch-verify \
  -s scrutiny/schemas/jcAlgSupport.yml \
  -r examples/jcAlgSupport/Infineon_CJTOP_80K_INF_SLJ_52GLA080AL_M8.4_ICFabDate_2012_001_ALGSUPPORT.json \
  --profiles-dir tests/mapper/test-data/jcAlgSupport \
  --profile-type jcalgsupport \
  --report-mode nonmatch

Documentation

For deeper usage and development information, see:

• CLI guide
• Schema overview
• Examples guide
• Troubleshooting
• Plugin module overview
• Test structure

Choosing the right command

• Use map when your input is raw and supported by a mapper.
• Use verify when you already have two normalized JSON files.
• Use report when you already have a verification JSON.
• Use full when you want one command for the entire single comparison pipeline.
• Use batch-verify when you want to compare one reference against many profiles.

For developers

The project is organized around modular:

• mappers for normalization
• comparators for section-level comparison
• viz plugins for report rendering
• schemas for structure, ingest behavior, comparator selection, and report types