mirror of
https://github.com/SHOGGOTH-SECTOR/sica-fondt.git
synced 2026-08-01 08:30:20 +00:00
24f816b6a3
Place useful parts of the surrounding repos into sica-fondt by layer, per the
body model (Ada = membrane; brain/endocrine/capabilities/knowledge non-Ada):
- brain/ LLM reasoning + providers (dapr, hermes, MoMoA)
- capabilities/ REPRAG sidecars: hermes tools/skills, dapr tools, parallel
dispatch, A51 channels, and the OSINT cluster
- knowledge/ LORAG corpus: 754 cyber-skills, agency personas, secure-coding,
MITRE ATT&CK data
- reference/ defensive threat-reference (C3, shhbruh doc) + AdaYaml parser
License handling: AGPL sources (worldosint, advanced_evolution, mercury,
Reticulum) and GPL DeTTECT are SPEC-only clean-room/port descriptions — no
copyleft code copied. MIT/Apache/data parts copied as working trees.
Safety: shhbruh escape/persistence material and C3 covert-C2 kept as reference
only, not wired into the running organism. See CONSOLIDATION.md.
https://claude.ai/code/session_01UehUqEXXJJCsHoA4voCU5c
93 lines
5.0 KiB
Markdown
93 lines
5.0 KiB
Markdown
# Workflow Reference: Secret Scanning with Gitleaks
|
|
|
|
## Secret Detection Pipeline
|
|
|
|
```
|
|
Developer Workstation CI/CD Pipeline Security Response
|
|
│ │ │
|
|
▼ │ │
|
|
┌──────────────┐ │ │
|
|
│ Pre-commit │ │ │
|
|
│ Hook (local) │ │ │
|
|
└──────┬───────┘ │ │
|
|
│ │ │
|
|
┌────┴────┐ │ │
|
|
│ │ │ │
|
|
PASS FAIL │ │
|
|
│ (blocked) │ │
|
|
▼ │ │
|
|
Push to │ │
|
|
Remote │ │
|
|
│ ▼ │
|
|
│ ┌──────────────┐ │
|
|
└───────────────────────>│ Gitleaks CI │ │
|
|
│ Scan (PR) │ │
|
|
└──────┬───────┘ │
|
|
│ │
|
|
┌─────┴─────┐ │
|
|
│ │ │
|
|
PASS FAIL │
|
|
│ ┌─────┴──────┐ │
|
|
│ │ Block PR │ │
|
|
│ │ + Alert │──────────────>│
|
|
│ └────────────┘ ┌──────────┴──────┐
|
|
│ │ Rotate Secret │
|
|
│ │ Update Baseline │
|
|
│ │ Clean History │
|
|
│ └─────────────────┘
|
|
▼
|
|
Merge Permitted
|
|
```
|
|
|
|
## Gitleaks Rule Configuration Deep Dive
|
|
|
|
### Rule Anatomy
|
|
```toml
|
|
[[rules]]
|
|
id = "rule-unique-identifier" # Unique rule ID
|
|
description = "Human-readable desc" # What this rule detects
|
|
regex = '''pattern''' # Detection regex
|
|
entropy = 3.5 # Minimum entropy threshold (optional)
|
|
secretGroup = 1 # Regex capture group containing secret
|
|
keywords = ["key1", "key2"] # Fast pre-filter keywords
|
|
tags = ["aws", "credential"] # Categorization tags
|
|
path = '''\.env$''' # Path filter regex (optional)
|
|
```
|
|
|
|
### Built-in Rule Categories
|
|
| Category | Example Rules | Count |
|
|
|----------|--------------|-------|
|
|
| Cloud Provider Keys | aws-access-key-id, gcp-service-account | 15+ |
|
|
| API Tokens | github-pat, gitlab-pat, slack-token | 20+ |
|
|
| Private Keys | private-key, rsa-private-key | 5+ |
|
|
| Database Credentials | generic-password, connection-string | 10+ |
|
|
| Service Tokens | stripe-api-key, sendgrid-api-key | 30+ |
|
|
|
|
### Entropy Scoring
|
|
- Entropy measures string randomness (Shannon entropy)
|
|
- Random-looking strings (API keys) have entropy > 3.5
|
|
- Regular English text has entropy around 2.0-3.0
|
|
- Setting entropy threshold reduces false positives on non-random strings
|
|
- Combine entropy with regex for highest accuracy
|
|
|
|
## Remediation Process
|
|
|
|
### Secret Rotation Checklist
|
|
1. Identify the exposed secret type and associated service
|
|
2. Log into the service provider and revoke the exposed credential
|
|
3. Generate a new credential with the same permissions
|
|
4. Store the new credential in a secrets manager (Vault, AWS SM, etc.)
|
|
5. Update all consuming services to use the new credential
|
|
6. Verify service functionality with the new credential
|
|
7. Update the Gitleaks baseline to remove the resolved finding
|
|
8. Optionally clean git history with git-filter-repo
|
|
|
|
### History Cleanup Decision Matrix
|
|
| Factor | Clean History | Keep History |
|
|
|--------|--------------|--------------|
|
|
| Secret is rotated | Optional | Acceptable |
|
|
| Repo is public | Required | Never |
|
|
| Compliance mandate | Required | Not compliant |
|
|
| Active contributor count | < 10 preferred | > 10 difficult |
|
|
| Secret exposure duration | Long (high risk) | Short (lower risk) |
|