Research: security-best-practices
Cached research evidence for security-best-practices (not authority).
Purpose
Section titled “Purpose”Security best practices for agents/apps. From official openai/skills catalog for Codex (and compatible). Curated or experimental skills in .curated/.experimental.
Harness Coverage
Section titled “Harness Coverage”Target agents: antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode.
Trust And Risks
Section titled “Trust And Risks”trust_tier=needs-inspection; status=inspect-then-install; provenance=verified-install-command. Web-aug: github.com/openai/skills (high stars ~22k, active OpenAI team, Python/JS/Shell, follows agentskills.io standard, documented install via $skill-installer). Inspect for broad scopes in security/CLI creator (write/exec potential). Catalog flags for inspection before promotion.
Install Prerequisites
Section titled “Install Prerequisites”Install: npx skills add openai/skills --skill security-best-practices -y -g -a antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode (or via specific .experimental etc) status=inspect-then-install; selector=named. Restart agent after install per docs.
Upstream Maintainer
Section titled “Upstream Maintainer”openai/skills (official OpenAI, high activity)
Comparable Alternatives
Section titled “Comparable Alternatives”Anthropics skills, local security-auditor / threat-model skills; other platform-specific (e.g. mcp-use).
> Web from repo README 2026 + catalog notes.