registry / @dujavi/ai-md / 0.6.0

@dujavi/ai-md@0.6.0

AXI-shaped CLI for private ~/.ai-md: shared+agents build to dist, multi-harness link (Cursor, Claude, agents, …)

AI Security Review

scanned 4h ago · by lpm-firewall-ai

Review flagged AI-agent configuration or capability changes. This remains warn-only unless evidence shows foreign-agent hijack through preinstall/install/postinstall, hidden persistence, exfiltration, remote code execution, or other concrete malicious behavior.

Static reason

No blocking static signals were detected.

Trigger

User runs `ai-md install`, `build`, `pull`, `link`, `harness`, or `script`.

Impact

A user-selected or configured private repository can populate local AI-agent rules/skills and run its own scripts; this is a meaningful agent-control capability but is not automatic.

Mechanism

Explicit AI-harness symlink/config management and user-private script execution.

Rationale

No concrete malicious chain was found, but the package has explicit user-invoked AI-agent control-surface writes and private-script execution. Per the firewall policy, this warrants a warning rather than a block.

Evidence

package.jsonbin/ai-md.jslib/commands.jslib/config.jslib/scripts.jslib/harnesses/index.jsscripts/link-project.sh~/.ai-md~/.ai-md/dist/<harness>~/.ai-md/scripts/<name>~/.cursor/skills~/.cursor/rules~/.claude/skills~/.claude/rules~/.agents/skills<repo>/.cursor<repo>/.gitignore

@dujavi/ai-md@0.6.0

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code