---
title: "Trail of Bits Agentic Actions Auditor: review, setup & limitations · Undominated.ai"
canonical: https://undominated.ai/skills/trailofbits-agentic-actions-auditor/
description: "Traces external issue or pull-request input into AI-enabled GitHub Actions and reports the permissions that could amplify an injection path."
---

# Trail of Bits Agentic Actions Auditor: review, setup & limitations · Undominated.ai

> Traces external issue or pull-request input into AI-enabled GitHub Actions and reports the permissions that could amplify an injection path.

[← Explore all skills](/skills/)

SECURITY / Trail of Bits

# Trail of Bits Agentic Actions Auditor

Traces external issue or pull-request input into AI-enabled GitHub Actions and reports the permissions that could amplify an injection path.

 [See setup guidance ↓](#setup)[Original source ↗](https://github.com/trailofbits/skills/tree/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/skills/agentic-actions-auditor)

SOURCE REVIEW

 Reviewed 2026-10-14
 Evidence 5 linked sources
 Publisher Trail of Bits
 Licence [CC-BY-SA-4.0 ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/LICENSE)
 Revision 8e54050ae63c
 [Read what was—and wasn’t—checked ↓](#review)

“Audits GitHub Actions workflows for security vulnerabilities in AI agent integrations including Claude Code Action, Gemini CLI, OpenAI Codex, and GitHub AI Inference.”

 [Trail of Bits · upstream description ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/skills/agentic-actions-auditor/SKILL.md) Our analysis follows below.

01 / THE REASONING

## Why this made the selection.

 - Checks environment intermediaries and reusable-workflow inputs as well as prompt text.
- Separates permissive configuration from a demonstrated attacker-input path instead of treating either alone as exploitation.

### A good fit for

 - Reviewing an AI pull-request reviewer before enabling external triggers
- Checking composite and reusable workflows for hidden prompt inputs
- Trace untrusted event data into CI agent prompts through reusable workflows and composite actions.

### Weigh up before choosing

 - Static guidance only: it does not execute an attack, fix workflows or certify that CI is secure.
- Cross-file resolution stops after one level; nested references remain unresolved, and generic remote actions and Docker actions are outside that traversal.
- Known action names and configuration fields can change; confirm the action revision before relying on a profile.
- Static data flow identifies risk, not a demonstrated exploit. An allowlisted tool or environment variable does not sanitize attacker-controlled content, and missing remote source is a review gap.

02 / THE REVIEW RECORD

## What we actually inspected.

Source review has boundaries. A clear record is more useful than a “safe” badge.

### Material inspected

 - plugins/agentic-actions-auditor/skills/agentic-actions-auditor/SKILL.md
- plugins/agentic-actions-auditor/skills/agentic-actions-auditor/references/cross-file-resolution.md
- plugins/agentic-actions-auditor/README.md
- README.md
- LICENSE

### Our findings

 - The definition records workflow triggers, environment data, tool/sandbox settings, permissions and action-specific prompt sinks.
- The reference distinguishes local composites from reusable workflows and carries caller-input mappings across the supported boundary.
- Remote review is limited to GitHub API reads; fetched YAML is treated as data, and unresolved references must constrain the coverage claim.
- The workflow explicitly follows env intermediaries and cross-file references and requires the chosen ref on every remote file fetch rather than only the directory listing.

### Not established by this review

 - No skill installation, consuming-agent workflow or external service operation was executed.
- Source inspection does not establish runtime compatibility, security certification or workload performance.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

## Add a skill to your workflow.

[Upstream setup instructions ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/README.md)
 - In a compatible plugin host, add the trailofbits/skills marketplace and select agentic-actions-auditor; inspect the complete plugin at the linked revision.
- Provide a local repository or an exact remote repository/ref. For private remote reads, configure GitHub CLI access with the required repository visibility.
- Review the report’s unresolved references and confirm reachable input-to-agent paths before treating any issue as exploitable.
- Pin every fetched workflow/action reference, inspect transitive uses targets and record trigger, token/secrets, sandbox and prompt inputs before assigning findings.

### Before you start

 - A compatible Claude Code or Codex plugin host
- GitHub workflow files; GitHub CLI and repository access for remote analysis

### Compatibility

Claude Code or Codex plugin host with the required tools

### Read workflow and referenced configuration files

 - Read workflow and referenced configuration files
- Use GitHub API reads for remote workflow resolution
- Produce findings without modifying or executing the reviewed workflows
- Reads workflow/action source using GitHub access; the defined review does not execute fetched YAML or perform live prompt-injection tests.

### Cost model

The CC-BY-SA-4.0 instructions are separate from agent usage and GitHub service entitlements.

04 / FOLLOW THE EVIDENCE

## The source trail.

Our notes are separate from the original resource. Check upstream before adopting a new version.

 - [Skill definition at reviewed revision ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/skills/agentic-actions-auditor/SKILL.md) Checked 2026-10-14 https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/skills/agentic-actions-auditor/SKILL.md Supports: summary, upstreamDescription, whySelected, bestFor, limitations, review, install, compatibility, access
- [Inspected source: plugins/agentic-actions-auditor/skills/agentic-actions-auditor/references/cross-file-resolution.md ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/skills/agentic-actions-auditor/references/cross-file-resolution.md) Checked 2026-10-14 https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/skills/agentic-actions-auditor/references/cross-file-resolution.md Supports: review, install, limitations, access
- [Inspected source: plugins/agentic-actions-auditor/README.md ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/README.md) Checked 2026-10-14 https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/plugins/agentic-actions-auditor/README.md Supports: review, install, limitations, access
- [Inspected source: README.md ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/README.md) Checked 2026-10-14 https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/README.md Supports: review, install, limitations, access
- [Inspected source: LICENSE ↗](https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/LICENSE) Checked 2026-10-14 https://github.com/trailofbits/skills/blob/8e54050ae63c2f62224c35b1616efe71865b4777/LICENSE Supports: license

KEEP COMPARING

## Other approaches to consider.

Related by category or shared topics. These are alternatives to inspect, not a measured quality order.

 [### Infisical User Setup Guide ↗ Infisical skill for getting secrets from Infisical into applications and platforms: the CLI, Docker, CI/CD, Kubernetes, the language SDKs and machine identity authentication.](/skills/infisical-setup/)[### Burp Project Parser ↗ Extract and search captured Burp traffic while distinguishing empty results from parser startup failures.](/skills/trailofbits-burpsuite-project-parser/)[### C and C++ Security Review ↗ Coordinate scoped native C/C++ security review and assemble findings with an explicit coverage ledger.](/skills/trailofbits-c-review/)

 [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)[Workflows ↗](/workflows/)

## Continue your investigation

 - [Choose a toolkit](/tools/)
- [Inspect execution hosts](/agents/)
- [Check tool access](/mcp-servers/)
- [Check project dependencies](/check-repo/)
