agentic-actions-auditor

agentic-actions-auditor18 wordsInspect firstinspect-then-install

Curated third-party skill source. Run external-skill-auditor before repo promotion.

Quick Start

Install:

npx skills add trailofbits/skills --skill differential-review --skill agentic-actions-auditor --skill variant-analysis --skill insecure-defaults --skill supply-chain-risk-auditor --skill modern-python -y -g -a antigravity claude-code codex crush cursor gemini-cli github-copilot grok opencode

Use: /agentic-actions-auditor

Works with Claude Code, Gemini CLI, OpenCode, and other agentskills.io-compatible agents.

Curated external (hand-maintained)

Composed compose-external-wave-1 using authoring SSOT (docs/src/authoring/skills/agentic-actions-auditor.mdx) + research. Do not run wagents docs generate on this file (HAND-MAINTAINED sentinel). Run external-skill-auditor before using in production or promoting.

What It Does

Audits GitHub Actions workflows for AI agent security vulnerabilities specific to integrations like Claude Code Action, Gemini CLI, OpenAI Codex, and GitHub AI Inference. Detects 9 attack vectors (env var intermediary, direct ${{}} injection, gh CLI data fetch at runtime, pull_request_target+checkout, error log injection, subshell bypass, eval of AI output, dangerous sandboxes like —yolo, wildcard allowlists). From Trail of Bits skills marketplace.

Harness Coverage

Targets verified harnesses: antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode.

Portable multi-harness install command:

npx skills add trailofbits/skills --skill differential-review --skill agentic-actions-auditor --skill variant-analysis --skill insecure-defaults --skill supply-chain-risk-auditor --skill modern-python -y -g -a antigravity claude-code codex crush cursor gemini-cli github-copilot grok opencode

Trust / Audit

Trust tier: Inspect first (needs-inspection)

Curated status: inspect-then-install

Risk notes: Curated third-party skill source. Run external-skill-auditor before repo promotion.

Entry maintained via authoring + research for compose-external-wave-1; provenance and audit notes are authoritative there (research context is advisory).

Install Prerequisites

Install: npx skills add trailofbits/skills --skill agentic-actions-auditor -y -g -a antigravity claude-code codex crush cursor gemini-cli github-copilot grok opencode status=inspect-then-install; selector=named. Best with .github/ workflows present. Run external-skill-auditor first per policy.

Public Metadata

Field	Value
Source Type	curated-external
Display Source	trailofbits/skills
Source Kind	github
Installability	portable command
Review State	curated
Trust Tier	needs-inspection
Target Agents	antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode

General

Field	Value
Name	agentic-actions-auditor

Curated catalog entry

docs/src/authoring/skills/agentic-actions-auditor.mdx (full SSOT excerpt)

---
name: "agentic-actions-auditor"
description: "Curated third-party skill source. Run external-skill-auditor before repo promotion."
title: "Agentic Actions Auditor"
source_kind: "curated-external"
source: "trailofbits/skills"
install_source: "trailofbits/skills"
status: "inspect-then-install"
trust_tier: "needs-inspection"
provenance_status: "verified-install-command"
install_command: "npx skills add trailofbits/skills --skill differential-review --skill agentic-actions-auditor --skill variant-analysis --skill insecure-defaults --skill supply-chain-risk-auditor --skill modern-python -y -g -a antigravity claude-code codex crush cursor gemini-cli github-copilot grok opencode"
target_agents: [antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode]
source_url: "https://github.com/trailofbits/skills"
notes: "Curated third-party skill source. Run external-skill-auditor before repo promotion."
risk_notes: "Curated third-party skill source. Run external-skill-auditor before repo promotion."
promotion_policy: "Inspect source, hooks, scripts, credentials, and dedupe before install."
provenance_evidence: "Curated `npx skills add` command with named `--skill` selectors under `inspect-then-install` in config/external-skills.md."
---

{/* GENERATED-AUTHORING: source=config/external-skills.md; entry=agentic-actions-auditor; re-run migration to refresh */}

Curated third-party skill source. Run external-skill-auditor before repo promotion.

Install / provenance (from authoring frontmatter + research):

Field	Value
install_command	npx skills add trailofbits/skills --skill differential-review --skill agentic-actions-auditor --skill variant-analysis --skill insecure-defaults --skill supply-chain-risk-auditor --skill modern-python -y -g -a antigravity claude-code codex crush cursor gemini-cli github-copilot grok opencode
source	trailofbits/skills
source_url	https://github.com/trailofbits/skills
trust_tier	needs-inspection
curated_status	inspect-then-install
target_agents	antigravity, claude-code, codex, crush, cursor, gemini-cli, github-copilot, grok, opencode

Resources

Skill Catalog Browse custom and external skills.

CLI Reference Install and manage skills.