Six installable knowledge packages that give AI agents real-world security and compliance expertise. Playbooks, evidence patterns, cross-framework leverage, board reporting, risk prioritization, MSSP program operation. Free, open source, vendor-neutral.
Skills follow the open Agent Skills standard — they activate automatically based on conversation context, work with any Agent Skills-compliant tool, and require no API key or runtime configuration.
No configuration. No API keys. The install script copies the skills into your local Claude Code skills directory; restart Claude Code to pick them up.
git clone https://github.com/BlueRadius360/radius360.gitcd radius360 && ./scripts/install.shTry one of:
"Walk me through a SOC 2 quarterly access review."
"What playbook do I run for a missing BAA?"
"Draft a 3-paragraph board briefing from these numbers."
"We have SOC 2 — what does that buy us toward HIPAA?"
RADIUS360_SKILLS_DIR=/path/to/skills before running the script.Each skill is a self-contained knowledge package. They activate automatically based on what you're asking — no prompting tricks, no manual selection.
Twelve canonical security and compliance playbooks. When to apply each, the steps, and the controls they satisfy across SOC 2, ISO 27001, HIPAA, PCI, NIST CSF, GDPR.
View on GitHubAuditor-grounded evidence patterns. What artifacts auditors actually accept per control, where they live, and the most common collection mistakes.
View on GitHubHow investment in one framework leverages work toward another. SOC 2 ↔ ISO 27001 ↔ HIPAA ↔ PCI ↔ NIST CSF ↔ GDPR.
View on GitHubBoard-ready security briefing structure, plain-language conventions, control-citation discipline, and the failure modes that turn briefings into status updates.
View on GitHubRisk scoring that survives audit. Impact × likelihood × asset × business-process methodology, percent-rank anchoring, and methodology versioning.
View on GitHubPatterns for running a portfolio of client security programs at vCISO/MSSP scale. Multi-tenant hygiene, baseline cascading, practitioner-scale delivery.
View on GitHubThe skills cite specific control IDs from the standards below — never invented controls, never marketing hand-waving. We do not redistribute the standards themselves; always reference the authoritative source for exact requirements.
Running portfolios of client security programs. Skills package the patterns that scale — playbook discipline, baseline cascading, portfolio rhythm.
Ground your AI assistant in real practitioner knowledge. Cite specific controls, propose specific evidence, write board-grade narrative without the hedge language.
Drop expert-grade GRC reasoning into your custom agents. Skills follow the open Agent Skills standard — works with any compliant SDK or tool.
Apache 2.0. No signup. No telemetry. The install script copies files locally and nothing else.
Skills are reference material for AI agents. They provide informational guidance and do not constitute legal or audit advice.