feat(S1): Windows Certificate Store scanning

[Rowuni]

Issue liée

close: #7

Description

Scanning du Windows Certificate Store et affichage de tout ce qui a été trouvé, classé suivant la sévérité. Le résultat de chaque scan se trouvera dans le fichier agent/scan_results.md, qui sera créé ou écrasé à chaque scan (fichier ajouté au gitignore)
Aussi, j'ai ajouté des catégories aux logs, comme [ERROR] ou [WARNING].

Pour tester :

cd agent
python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt
python scan.py

Puis regarder le fichier scan_results.md

Changements

Backend

• Ajout de la fonctionnalité de scan du Windows Certificate Store
• Ajout des catégories aux logs

Frontend

• Aucun

Checklist

• tsc --noEmit passe sans erreur
• Le backend démarre (uvicorn main:app)
• Pas de secret / clé / token commité
• Les types TypeScript sont alignés sur les modèles Pydantic
• Les changements nécéssaires à la documentation ont été faits

[Copilot]

Pull request overview

This PR implements the S1 scanner for the Windows Certificate Store and adds a generated Markdown report (agent/scan_results.md) to present scan findings grouped by severity, alongside categorized logging output.

Changes:

• Add a Windows Certificate Store scanner (wincertstore + cryptography) that parses certs, computes severity, and emits actionable recommendations.
• Generate a Markdown report (scan_results.md) after each run summarizing findings by scanner and severity.
• Update logging to include level categories (e.g., [INFO], [WARNING]), and ignore the generated report via .gitignore.

Reviewed changes

Copilot reviewed 3 out of 4 changed files in this pull request and generated 2 comments.

File	Description
agent/scanners/cert_store.py	Implements Windows cert store enumeration/parsing with severity, details, and recommendations.
agent/scanners/__init__.py	Documents a cross-scanner convention for algorithm label formatting.
agent/scan.py	Adds Markdown report generation and logging formatting; aggregates scanner results.
.gitignore	Ignores the generated scan_results.md report file.

Comments suppressed due to low confidence (1)

agent/scan.py:83

• The sorting key treats severity=None as unknown (falls back to 99), so findings that explicitly have severity: null will be pushed to the end instead of being treated as "info". Consider normalizing with x.get('severity') or 'info' before looking up _SEV_ORDER.

        for f in sorted(findings, key=lambda x: _SEV_ORDER.get(x.get("severity", "info"), 99)):
            sev = f.get("severity", "info")
            algo = f.get("algorithm") or "—"
            store = f.get("store_or_path") or f.get("key_path") or f.get("path") or "—"
            name = str(f.get("name") or "—")[:50]

[GorgorQ]

Approved !