AI skillAI Coding Agents
AI Coding Agents AI security skills
Antigravity CLI, GitHub Copilot CLI, Claude Code, and OpenAI Codex workflows for secure review, remediation, CI/CD audit, and agent governance.
Security tool skills
AI skills for the security tools teams already use.
Practical, evidence-first workflows for using AI with SIEM, CNAPP, endpoint, identity, AppSec, vulnerability management, and model gateway tools without losing control. Now includes secure coding-agent workflows for Antigravity CLI, GitHub Copilot CLI, Claude Code, and OpenAI Codex.
AI skillSplunk
Splunk AI security skills
Alert triage, SPL drafting, detection tuning, incident timelines, and SOC handoffs.
AI skillWiz
Wiz AI security skills
Attack path summaries, exposure prioritization, cloud ownership, and remediation planning.
AI skillPalo Alto Networks
Palo Alto Networks AI security skills
Firewall policy review, Prisma Cloud risk summaries, Cortex investigations, and change review.
AI skillCrowdStrike
CrowdStrike AI security skills
Endpoint process tree summaries, containment notes, threat hunting, and incident updates.
AI skillMicrosoft Sentinel
Microsoft Sentinel AI security skills
KQL query drafting, incident summaries, Microsoft 365 investigations, and automation review.
AI skillOkta
Okta AI security skills
Suspicious login review, policy analysis, access review, identity incident recovery, and blast-radius summaries.
AI skillGitHub
GitHub AI security skills
Pull request security review, code scanning triage, secret response, dependency review, and CI/CD checks.
AI skillTrivy
Trivy AI security skills
Vulnerability clustering, container triage, SBOM review, IaC remediation, and risk briefs.
AI skillLiteLLM
LiteLLM AI security skills
Model gateway policy review, virtual key control, token budgets, prompt log triage, and provider routing.
Coding agents
Use the right AI coding assistant for the right security job.
The safest pattern is not one magic assistant. It is a governed toolchain where each coding agent has a clear job, bounded access, auditable output, and human approval for risky changes.
Coding agentSecurity workflow
OpenAI Codex
Repository-scale implementation, secure refactors, test generation, and evidence-backed code changes.
Guardrail
Run in a branch, require diffs and tests, restrict secrets, and keep deployment behind human approval.
Coding agentSecurity workflow
Claude Code
Security review, threat-model drafting, dependency triage, and explaining complex code paths.
Guardrail
Keep skills versioned, review tool permissions, and avoid broad filesystem or shell access for untrusted workflows.
Coding agentSecurity workflow
GitHub Copilot
Developer-in-IDE secure coding support, pull request assistance, tests, and lightweight remediation guidance.
Guardrail
Pair with branch protection, code scanning, secret scanning, dependency review, and required reviewer policies.
Coding agentSecurity workflow
Antigravity CLI
Command-line AI-assisted investigation, repetitive security checks, local developer workflows, and scripted triage.
Guardrail
Scope shell commands, log prompts and outputs, avoid production credentials, and prefer read-only investigation by default.
Operating model
Use AI to compress evidence, not replace evidence.
Keep the tool as system of record
Splunk events, Wiz graph paths, CrowdStrike detections, Okta logs, GitHub alerts, and Trivy findings should remain traceable in every AI-generated summary.
Require approval for action
Let AI draft queries, tickets, summaries, and remediation plans. Keep containment, policy changes, identity changes, and production deploys behind human review.