@cyberstrategyinstitute/ai-safe2-framework

> This file (root `skill.md`) covered AI SAFE2 v2.1 (128 controls, 14 frameworks).

skillclaude

Install

agr install @cyberstrategyinstitute/ai-safe2-framework --target claude

Writes 19 files into .claude/skills/, pinned to git-cda649d3.

.claude/skills/ai-safe2-framework/.markdownlint.yaml
.claude/skills/ai-safe2-framework/ADVANCED_AGENT_THREATS.md
.claude/skills/ai-safe2-framework/CITATION.cff
.claude/skills/ai-safe2-framework/CODEOWNERS
.claude/skills/ai-safe2-framework/CODE_OF_CONDUCT.md
.claude/skills/ai-safe2-framework/CONTRIBUTING.md
.claude/skills/ai-safe2-framework/CSF-COMPANION.md
.claude/skills/ai-safe2-framework/Dockerfile
.claude/skills/ai-safe2-framework/EVOLUTION.md
.claude/skills/ai-safe2-framework/INTEGRATIONS.md
.claude/skills/ai-safe2-framework/LICENSE
.claude/skills/ai-safe2-framework/MAINTAINERS.md
.claude/skills/ai-safe2-framework/QUICKSTART_5_MIN.md
.claude/skills/ai-safe2-framework/README.md
.claude/skills/ai-safe2-framework/SECURITY.md
.claude/skills/ai-safe2-framework/VANGUARD_PROGRAM.md
.claude/skills/ai-safe2-framework/docker-compose.yml
.claude/skills/ai-safe2-framework/pyproject.toml
.claude/skills/ai-safe2-framework/skill.md

Document

AI SAFE2 Secure Build Copilot — v3.0

This file (root skill.md) covered AI SAFE2 v2.1 (128 controls, 14 frameworks). It has been superseded by the v3.0 skill ecosystem in the skills/ folder.

For Claude Projects / Claude Desktop

Upload skills/SKILL.md to your Project Knowledge instead of this file.

AI SAFE2 v3.0: 161 controls, 32 frameworks, CP.1-CP.10 cross-pillar governance, HEAR Doctrine, Agent Replication Governance, AIVSS AAF scoring integration.

For Live Tool Access (Claude Code, Codex)

See skills/mcp/README.md for the MCP server setup guide. Local stdio: 5 minutes, no token. Remote HTTPS: 15 minutes via Railway.

What Changed in v3.0

	v2.1 (this file)	v3.0 (skills/)
Controls	128	161 (151 pillar + 10 CP)
Frameworks	14	32
MCP server	Referenced only	Fully built
HEAR Doctrine	Not defined	CP.10 (first in field)
Agent Replication	Not defined	CP.9 (first in field)
Risk formula	CVSS + Pillar	CVSS + Pillar + AAF

Trustgrade B

passBody integrity
Whether the stored document is plausibly the kind of file the artifact declares, rather than something fetched by mistake.
passType matchnot applicable to this artifact type
Whether the artifact is really the kind of thing its metadata claims it is.
passFreshness
How long since the source repository was last pushed to.
passPrompt injection
Scans the artifact's own text for instructions aimed at your agent rather than at you.
warnLicenseno SPDX license detected
Whether the source repository declares an SPDX license permissive enough to redistribute.

How the grade is calculated

Each check contributes 0 points when it passes, 1 when it warns, and 2 when it fails. The total maps to a letter:

Aevery check passed
Bone warning
Ctwo warnings
Dprompt injection or body integrity failed, or three warnings
Fone of those failed, and something else is wrong

These are automated hygiene checks, not a security audit, and not a dependency or vulnerability scan. A grade of A means nothing was flagged — not that the artifact is safe.

Versions

git-cda649d331ce2026-07-31