←Leaderboard/Security/hack-skills
Last commit on September 13, 2026·Created on April 7, 2026

yaklang/hack-skills

“A library of expert attack playbooks for penetration testers and security researchers.”
Combined rank
#192
across all skills
In Security
#10
category rank
Stars
2.5k
+2.7% in last 7d
Forks
337
+3.7% in last 7d
Watchers
13
0.0% in last 7d
Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
TL;DR

This collection provides structured technical guidance for executing advanced attack vectors across Active Directory, API authorizations, and security control bypasses. It translates complex exploitation paths into actionable playbooks for professional security assessments.

Penetration testers can use these guides to systematically approach AD ACL abuse, Kerberos attacks, and the circumvention of anti-debugging measures.

WHO IT'S FOR
Penetration testers
executing advanced attack playbooks for AD and APIs
Red team operators
bypassing security controls and anti-debugging measures
AI agent / automation builders
integrating hacking capabilities into AI agents
Android security researchers
performing deep analysis of Android applications
Categories
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.

Use this skill collection

Follow the documented setup, then try a first task.

Install the complete bundle with npx

Installs: Complete HackSkills bundle: one master entry, six category entries, and the repository's deep-topic skills.

Before you start
  • The npx command must be available.
  1. Run this terminal command to add the complete yaklang/hack-skills bundle.

    npx skills add yaklang/hack-skills
skills/injection-checking/SKILL.md ↗ · Checked Oct 18, 2026skills/api-sec/SKILL.md ↗ · Checked Oct 18, 2026skills/auth-sec/SKILL.md ↗ · Checked Oct 18, 2026skills/recon-for-sec/SKILL.md ↗ · Checked Oct 18, 2026README.md ↗ · Checked Oct 18, 2026

Give it something to do.

Suggested first task

Build a first-pass API surface inventory

Uses api-recon-and-docs

  • Authorized API base URL
  • Client JavaScript files or URLs, if available
  • OpenAPI, Swagger, or other API documentation, if available
  • In-scope hosts and testing boundaries
Use the api-recon-and-docs skill to perform a first-pass review of my authorized API target. From the supplied base URL, client JavaScript, and any OpenAPI or Swagger documents, produce a compact inventory of reachable entry points, versions, schemas, optional fields, role differences, GraphQL endpoints, mobile paths, and undocumented parameters. Identify the three highest-value areas for follow-up testing without attempting exploitation.
Suggested first task

Create a focused two-account authorization test matrix

Uses api-authorization-and-bola

  • Authorized testing scope
  • Sanitized request and response captures for Account A
  • Sanitized request and response captures for Account B
  • Known object identifiers and role differences
Use the api-authorization-and-bola skill to review the supplied authorized API request captures for Account A and Account B. Create a small replay matrix covering create, read, update, and delete operations, cross-account object identifiers, nested resources, alternate HTTP verbs, and hidden writable fields. Do not execute requests; return the prioritized matrix and the evidence each test should capture.
Suggested first task

Plan a benign CSV formula-evaluation check

Uses csv-formula-injection

  • Path replacing <sample-export-path>
  • Spreadsheet application and version
  • Authorized export workflow
  • List of user-controlled fields represented in the export
Use the csv-formula-injection skill to design a minimal, authorized check of the sample export at <sample-export-path>. Start only with benign arithmetic cells such as =1+1 and the documented prefix variants. Explain which user-controlled fields and export sinks to inspect, how to open the file in the named spreadsheet application, and how to record whether each cell remains text, calculates, or raises a warning. Do not use DDE, commands, remote URLs, or external-data functions.