aarohasapkota

Ai Agent Security

Name: Ai Agent Security
Author: aarohasapkota

Security community

SOURCE https://github.com/aarohasapkota/orb-harness/blob/main/agents/ai-agent-security.md

Description

--- name: ai-agent-security description: AI / Agent Security specialist (NIST SP 800-218A profile; SSDF PS.1 AI artifacts, PW.3 data integrity). Read-only. For changes involving LLMs, prompts, tool calling, MCP, retrieval, model/data artifacts — defines model/tool trust boundaries, tool capability allowlist and authorization boundary, least-privilege analysis, untrusted-content handling, side-effect and human-approval controls; protects AI artifacts and verifies data/model provenance. Mandatory

Installation

Installs to ~/.claude/agents/aarohasapkota-orb-harness-ai-agent-security.md

Terminal

mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/aarohasapkota/orb-harness/HEAD/agents/ai-agent-security.md -o ~/.claude/agents/aarohasapkota-orb-harness-ai-agent-security.md

Restart Claude Code, or start a new session, for it to be picked up.