Is the Agentic Actions Auditor skill free?

Yes. The skill itself is free from trailofbits/skills. SkillProof publishes the install command and an independent test verdict at no cost.

What is the Agentic Actions Auditor SkillProof Score?

9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 9/10, docs & honesty 4/5.

How do I install Agentic Actions Auditor?

Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.

Can I use Agentic Actions Auditor with Cursor, Copilot, Gemini CLI, Codex or other AI tools?

The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.

Agentic Actions Auditor

Static-audits GitHub Actions workflows for prompt-injection paths into Claude/Gemini/Codex CI agents

by trailofbits · trailofbits/skills

Tested · Works

Install — copy the command Download source ↗

Test report

Verdict: Tested · Works
Score: 9.2/10
Tested: Jul 10, 2026
Environment: Claude Code 2.x (agent harness)
Upstream re-checked: Jul 18, 2026 · 3b0255a

Given a workflow with wildcard permissions and env-var-sourced issue content, an unguided review caught the obvious flags but missed how the env var and prompt actually connect; the skill traced the full chain.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

Installs cleanly 5/5
Triggers reliably 5/5
Output vs. baseline 9/10
Docs & honesty 4/5

What Agentic Actions Auditor does

Walks through a 5-step methodology to find AI action steps (Claude Code Action, Gemini CLI, OpenAI Codex, GitHub AI Inference) in GitHub Actions workflows, trace whether attacker-controlled event data (issue/PR body, comments) reaches the agent's prompt, and grade findings by severity with remediation. Triggers when reviewing workflow files that invoke AI coding agents or auditing CI/CD pipelines for prompt-injection risk.

How to install Agentic Actions Auditor

git clone https://github.com/trailofbits/skills
cd skills
mkdir -p ~/.claude/skills
cp -r plugins/agentic-actions-auditor/skills/agentic-actions-auditor ~/.claude/skills/agentic-actions-auditor

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Agentic Actions Auditor

/agentic-actions-auditor Static-audits GitHub Actions workflows for prompt-injection paths into Claude/Gemini/Codex CI agents

It also activates on plain-language prompts like these:

Audit this GitHub Actions workflow for prompt-injection into Claude
Check if PR comment text can reach our Gemini CLI agent's prompt
Grade the severity of AI-agent prompt-injection risk in this CI pipeline

Frequently asked questions

Is the Agentic Actions Auditor skill free?: Yes. The skill itself is free from trailofbits/skills. SkillProof publishes the install command and an independent test verdict at no cost.
Does Agentic Actions Auditor work with Claude Code?: We tested it with Claude Code 2.x (agent harness) on Jul 10, 2026. Verdict: Tested · Works. Given a workflow with wildcard permissions and env-var-sourced issue content, an unguided review caught the obvious flags but missed how the env var and prompt actually connect; the skill traced the full chain.
What is the Agentic Actions Auditor SkillProof Score?: 9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 9/10, docs & honesty 4/5.
How do I install Agentic Actions Auditor?: Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Agentic Actions Auditor with Cursor, Copilot, Gemini CLI, Codex or other AI tools?: The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.