Threat-native blockchain security scanner + ecological intelligence platform. 108 blocks on chain. East Africa pilot active.
| Component | What is built and running |
|---|---|
| Trinity consensus mechanism | 2-node canonical voting that certifies each finding before it enters the chain, enforcing deterministic agreement between nodes. |
| OracleBrain | 8-layer adaptive neural threat classifier used for higher-order threat interpretation and adaptive scoring behavior. |
| ThreatNet | 22,083-parameter neural network classifier that performs fast risk classification on scanner findings and candidate attack paths. |
| COVENANT | Constitutional CI/CD guard that verifies chain integrity, applies policy thresholds, and can halt deployments on critical certified threats. |
| Multi-language scanners | Python, JavaScript, and Go scanners operating in one pipeline to detect taint flows, insecure patterns, and vulnerability signatures. |
| EDEN ecological intelligence | NDVI + acoustic sensor ingestion mapped into ecological-risk intelligence and BioImpact-linked certification context. |
| Unified CLI interface | Single control plane through mistcoder commands (scan, chain, brain, covenant, status, selftest) for end-to-end operation. |
| Dashboard (Editorial SVG frontend) | Visual reporting layer that renders certified chain and intelligence outputs into editorial SVG views for operators and stakeholders. |
- 108 verified blocks on-chain (integrity verified)
- 31 ecological events certified
- 279,787 tCO2 at risk tracked by ecological impact metrics
- 64 code vulnerabilities detected across current scan corpus
- East Africa pilot locations: Mau Forest, Rift Valley, Nairobi Eastlands
MISTCODER is built as a modular pipeline so each intelligence stage can evolve independently while sharing certified outputs:
- MOD-01 — Intermediate Representation (IR) generation
Converts source and findings into normalized structures for downstream analysis and correlation. - MOD-02 — Static & Dynamic Analysis Engine
Runs language-aware analysis flows, surfaces vulnerabilities, and prepares threat evidence for certification. - MOD-03 — Threat modeling and reasoning
Connects findings into attack-path logic, prioritizes adversarial pathways, and supports explainable decisioning. - Scoring + CVSS 3.1 integration
Combines model confidence, attack context, and CVSS-aligned severity to produce defensible and comparable risk scores.
- Multi-language threat detection for Python, JavaScript, and Go
- Blockchain-certified findings with auditable chain verification
- Ecological impact assessment tied to NDVI and acoustic telemetry
- Constitutional compliance checking via COVENANT policy gates
- Real-time threat classification with OracleBrain + ThreatNet
# Install MISTCODER in editable mode for local development and CLI usage
pip install -e .
# Run a multi-language security scan (Python, JavaScript, Go) on the current project
mistcoder scan . --lang py,js,go
# Check engine availability and platform readiness across all modules
mistcoder status
# Run built-in self-tests for key engines before CI/CD or pilot deployment
mistcoder selftest
# Verify COVENANT chain status and constitutional audit health
mistcoder covenant status
# Scan all configured EDEN pilot zones for ecological intelligence updates
python eden_cli.py scan --all-pilots- 108 blocks · integrity VERIFIED
- 31 ecological events certified · 279,787 tCO2 at risk
- 64 code vulnerabilities · Python + JavaScript + Go
- East Africa pilots: Mau Forest · Rift Valley · Nairobi Eastlands