Currently, techniques are only listed in Markdown under each tactic. This makes it difficult for developers and security teams to programmatically consume SAFE-MCP content.
Proposal:
- Add a machine-readable index (
techniques.json or techniques.yaml) with fields:
{
"id": "SAFE-T1102",
"name": "Prompt Injection",
"tactic": "Execution",
"description": "...",
"mitre_mappings": ["T1565.001"],
"mitigations": ["X", "Y"],
"detections": ["example"]
}
• Store under /data/ or /schemas/ directory.
• Add a GitHub Action to validate that new techniques added in Markdown are also updated in the JSON/YAML.
Why:
• Enables external tooling, dashboards, and automation to ingest SAFE-MCP techniques directly.
• Makes it easier to cross-link tactics ↔ techniques ↔ mitigations in docs or a future website.
Next step:
I can draft the schema and initial converter script to populate the JSON index from existing Markdown techniques.
Currently, techniques are only listed in Markdown under each tactic. This makes it difficult for developers and security teams to programmatically consume SAFE-MCP content.
Proposal:
techniques.jsonortechniques.yaml) with fields:{ "id": "SAFE-T1102", "name": "Prompt Injection", "tactic": "Execution", "description": "...", "mitre_mappings": ["T1565.001"], "mitigations": ["X", "Y"], "detections": ["example"] }• Store under /data/ or /schemas/ directory.
• Add a GitHub Action to validate that new techniques added in Markdown are also updated in the JSON/YAML.
Why:
• Enables external tooling, dashboards, and automation to ingest SAFE-MCP techniques directly.
• Makes it easier to cross-link tactics ↔ techniques ↔ mitigations in docs or a future website.
Next step:
I can draft the schema and initial converter script to populate the JSON index from existing Markdown techniques.