Skills github-workflow-automation

📦

github-workflow-automation

Name: github-workflow-automation
Author: sickn33

Content revision r2 High Risk ⚡ Contains scripts⚙️ External commands🌐 Network access📁 Filesystem access🔑 Env variables

Automate GitHub Workflows with AI

Manual reviews, triage, and repository maintenance consume engineering time. This skill provides adaptable GitHub Actions and TypeScript patterns for AI-assisted automation.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "github-workflow-automation" from https://skillstore.io/skills/sickn33-github-workflow-automation.md and its manifest at https://skillstore.io/api/skills/sickn33-github-workflow-automation/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/sickn33-github-workflow-automation.md Signed manifest GET /api/skills/sickn33-github-workflow-automation/manifest Signed lockfile GET /api/skills/sickn33-github-workflow-automation/lockfile

Test it

Using "github-workflow-automation". Review an AI pull request workflow before deployment.

Expected outcome:

Restrict the job to trusted events and grant only required permissions.
Pass pull request content as data, redact secrets, and validate AI responses.
Pin every action to a reviewed commit SHA.

Using "github-workflow-automation". Create issue triage for a small open-source repository.

Expected outcome:

Trigger on new issues and classify only against approved labels.
Request reproduction details when reports lack required information.
Require human review before assigning sensitive labels or closing issues.

Security Audit

High Risk

v5 • 7/24/2026 Open versioned report

Most static matches are Markdown syntax or standard secret wiring. Confirmed risks include dynamic shell commands, expression injection, unsafe output handling, untrusted AI prompts, external data disclosure, and unauthorized rebasing.

Files scanned

855

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (6)

High

SKILL.md:40-42 SKILL.md:77-82

Pull request data can inject JavaScript

Changed files and diff output are inserted directly into github-script source, allowing repository content to alter privileged JavaScript.

The workflow uses direct expression interpolation inside a JavaScript template literal while granting pull-request write permission.

High

SKILL.md:154-165

Changed filenames can inject shell syntax

PR-controlled changed-file output is inserted directly into a shell loop, so crafted filenames can become executable shell syntax.

The step embeds a previous output into shell source instead of passing it through a quoted environment variable.

High

SKILL.md:77-100 SKILL.md:197-223 SKILL.md:412-430 SKILL.md:715-727

Untrusted content controls AI decisions

Diffs, issue text, commit logs, and comments enter AI prompts without trust boundaries before outputs drive reviews, labels, checks, or comments.

Each example places repository-controlled text in an AI prompt and consumes the model response in an operational workflow.

High

SKILL.md:65-102

Repository content is sent to an external AI service

The review workflow sends complete pull request diffs to Anthropic without redaction, content limits, or repository data policy checks.

The code constructs an Anthropic request containing changed filenames and the complete diff.

High

SKILL.md:510-542

Any commenter can request a privileged rebase

The rebase job checks only comment text, then uses a repository token to rebase and force-push without author authorization.

No actor, association, or permission check appears between the public comment trigger and the write-capable Git commands.

Medium

SKILL.md:45 SKILL.md:66 SKILL.md:281 SKILL.md:483

Actions use mutable version tags

Workflow dependencies use movable major tags instead of reviewed commit SHAs, increasing supply-chain exposure if an upstream tag changes.

The uses entries visibly reference v4, v7, v9, and v1 tags rather than immutable commit identifiers.

Capability review items (21)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

High

SKILL.md:562

Python exec() function

const commitInfo = await exec(`git show ${commitHash} --stat`);

This TypeScript example passes interpolated refs and file data to a shell-backed exec function, enabling command injection.

High

SKILL.md:565

Python exec() function

const targetDiff = await exec(

This TypeScript example passes interpolated refs and file data to a shell-backed exec function, enabling command injection.

High

SKILL.md:583

Python exec() function

await exec(

This TypeScript example passes interpolated refs and file data to a shell-backed exec function, enabling command injection.

High

SKILL.md:586

Python exec() function

const result = await exec(`git cherry-pick ${commitHash}`, {

This TypeScript example passes interpolated refs and file data to a shell-backed exec function, enabling command injection.

High

SKILL.md:599

Python exec() function

await exec(`git checkout ${targetBranch}`);

This TypeScript example passes interpolated refs and file data to a shell-backed exec function, enabling command injection.

High

SKILL.md:600

Python exec() function

await exec(`git cherry-pick ${commitHash}`);

This TypeScript example passes interpolated refs and file data to a shell-backed exec function, enabling command injection.

High

SKILL.md:562

Process exec

const commitInfo = await exec(`git show ${commitHash} --stat`);

The exec call builds a shell command from unchecked commit, branch, or file values instead of using fixed argument arrays.

High

SKILL.md:565-567

Process exec

const targetDiff = await exec(

The exec call builds a shell command from unchecked commit, branch, or file values instead of using fixed argument arrays.

High

SKILL.md:583-584

Process exec

await exec(

The exec call builds a shell command from unchecked commit, branch, or file values instead of using fixed argument arrays.

High

SKILL.md:586-588

Process exec

const result = await exec(`git cherry-pick ${commitHash}`, {

The exec call builds a shell command from unchecked commit, branch, or file values instead of using fixed argument arrays.

High

SKILL.md:599

Process exec

await exec(`git checkout ${targetBranch}`);

The exec call builds a shell command from unchecked commit, branch, or file values instead of using fixed argument arrays.

High

SKILL.md:600

Process exec

await exec(`git cherry-pick ${commitHash}`);

The exec call builds a shell command from unchecked commit, branch, or file values instead of using fixed argument arrays.

High

SKILL.md:562-566

Ruby/shell backtick execution

const commitInfo = await exec(`git show ${commitHash} --stat`);