Guide · Agent skill
How to use the DPAPI Credential Chain agent skill (CTF)
Updated · written with AI from the project's own documentation and checked against our catalogue data
competition-dpapi-credential-chain is an agent skill: a set of written instructions (a SKILL.md file) that tells an AI agent how to work through one narrow task. This one covers capture-the-flag style exercises on Windows where the key secret is protected by DPAPI, Windows' built-in data-protection encryption. This guide explains what the skill makes the agent do, how to install it, how to trigger it, and what it needs before you use it.
DPAPI Credential Chain (CTF)zhaoxuya520/reverse-skill · MIT Download the skill Quick start
- 1Confirm your lab and scope. Make sure you have a sanctioned Windows CTF sandbox or lab, and that you are authorised to analyse its artifacts.
- 2Install the parent skill. Install and activate ctf-sandbox-orchestrator, and let it set the sandbox assumptions first.
- 3Download this skill's ZIP. Download the ZIP for competition-dpapi-credential-chain from this page.
- 4Unzip into a skills folder. Unzip into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder for one project.
- 5Or upload in the Claude apps. Alternatively, go to Settings → Capabilities → Skills and upload the skill there.
- 6Ask a scoped question. Describe the protected blob, masterkey or store in your lab and ask for the chain to the accepting service.
- 7Review the chain. Check the agent's output for source, unwrap and acceptance evidence before drawing conclusions.
What the skill makes the agent do
When active, the agent maps DPAPI-protected secrets to the masterkey and the user or machine context that can unwrap them. It then checks where the recovered plaintext is accepted, such as a browser session, SMB, RDP, WinRM or an app login. The guidance is a workflow and checklist: separate the protected blob, masterkey, decrypting context and final accepting service, and keep those links in one chain.
The skill is a downstream specialist. It is meant to run only after the parent skill ctf-sandbox-orchestrator has set sandbox assumptions and routed the task here. If it has not, the skill tells the agent to return to the parent first.
- Distinguish local user DPAPI, machine DPAPI, domain backup key use and application-specific wrapping.
- Record SID, profile path, masterkey path and credential store names.
- Separate a successful blob decryption from actual accepted access.
- Hand broader LSASS or full Windows pivoting work to a tighter host or pivot skill.
Install the skill from the ZIP
Download the ZIP for this skill from this page. The skill folder contains SKILL.md, references/dpapi-credential-chain.md and agents/openai.yaml. Then choose one of two install routes.
For a personal install, unzip the folder into ~/.claude/skills/. For a single project, unzip it into that project's .claude/skills/ folder. Alternatively, in the Claude apps, open Settings, go to Capabilities, then Skills, and upload the skill there.
- Personal: ~/.claude/skills/ (available across your projects).
- Project: .claude/skills/ inside the project folder (available only there).
- Claude apps: Settings → Capabilities → Skills, then upload.
Install the parent skill first
This skill will not work as intended on its own. The parent skill ctf-sandbox-orchestrator must be installed and active, and it must have established your sandbox assumptions before this one is used. Install and confirm the parent before you ask for DPAPI work.
The skill also needs a Windows CTF sandbox or lab with the artifacts you are analysing. It does not set up that environment for you.
- Parent skill ctf-sandbox-orchestrator installed and active.
- A Windows CTF sandbox or lab containing the artifacts.
- An agent runtime that supports SKILL.md skills.
Trigger the skill with natural requests
You do not need a special command. Describe the task in plain language that matches the skill's stated scope. The skill's description lists the kinds of requests it covers: inspecting DPAPI blobs or masterkeys, recovering browser or vault credentials, tracing DPAPI context or backup-key use, and explaining how protected Windows secrets become accepted access or privilege.
State up front that you are working in a sanctioned lab, and mention the parent skill so the agent knows the sandbox has been set up. The skill asks the agent to reply in Simplified Chinese unless you explicitly request English, so say so if you want English output.
- Inspect the DPAPI masterkey referenced by this lab's protected blob, after ctf-sandbox-orchestrator has set the sandbox assumptions.
- Trace which context unwraps the vault credential in this CTF image and where the plaintext is accepted.
- Explain how the protected Windows secrets in this lab become accepted access, in English please.
- Trace the backup-key use behind this DPAPI blob in the lab.
What the agent should produce
The skill asks for a minimal chain: protected artifact, then masterkey or unwrap context, then plaintext secret, then accepted replay or access, then the resulting capability. It also asks the agent to record the blob path, masterkey path, SID, protector scope, store name and the exact accepting service or dataset.
The reference checklist asks for evidence on three sides: the source (blob, store, SID, profile, scope), the unwrap (masterkey source, backup key use, plaintext type) and the acceptance (target service, host, unlocked data, resulting session or privilege).
- One minimal protected-artifact-to-accepted-access sequence that proves the edge.
- Source, unwrap and acceptance evidence grouped together.
- A clear statement of whether the decisive edge is masterkey recovery, DPAPI scope confusion, application wrapper handling or the accepting service.
Safety, scope and network notes
This skill is for sanctioned CTF or lab exercises. Its stated purpose includes recovering browser and vault credentials and replaying them, which is unlawful against systems you do not own or have written authorisation to test. Use it only in an isolated sandbox you control, with artifacts you are permitted to analyse.
According to the project's assessment, the skill runs no scripts and makes no network calls. It is a set of instructions for the agent, so the work it describes is carried out by your agent and tools in your environment. Keep your lab isolated regardless.
- Use only on systems you own or have written permission to test.
- Do not use it for general Windows administration or for recovering your own passwords; standard Windows tools or vendor documentation are more suitable there.
- The skill does not set up a safe sandbox. Users with no lab should not start here.
- The skill runs no scripts and uses no network, per the project's assessment.
Tips for better results
Give the agent the lab context early: which artifacts you have, which Windows user or machine they belong to, and what the flag or acceptance target is. The skill's first step depends on knowing the protected blob, the masterkey and the store.
Ask for the smallest proof rather than a broad sweep. The workflow is built around one reproducible path, and it tells the agent to hand off when the task widens.
- Name the artifacts and store (Credential Manager, Vault, browser Login Data, Wi-Fi profiles, RDP files or app storage).
- Ask for the accepting service explicitly, not just a decrypted value.
- Keep user and machine scope separate in your notes, as the checklist warns against mixing them.
What people use it for
- Tracing which user context can unwrap a DPAPI-protected credential in a training Windows image.
- Mapping the masterkey and profile path behind a browser Login Data store in a sanctioned lab.
- Explaining to a student how a protected Windows secret becomes an accepted login in a controlled exercise.
- Reviewing a red-team trainer's Windows credential-chain lab to confirm each link in the chain is documented.
Frequently asked questions
Is the DPAPI Credential Chain agent skill free?
This page does not state a price or licence terms for the skill itself beyond the project's listed licence, which is MIT. Check the licence file in the repository before reuse. The skill also depends on the parent skill ctf-sandbox-orchestrator, which you need to install separately.
What do I need to run this skill?
You need the parent skill ctf-sandbox-orchestrator installed and active, a Windows CTF sandbox or lab with the artifacts you are analysing, and an agent runtime that supports SKILL.md skills. The skill runs no scripts and uses no network, according to the project's assessment.
Where do I put the skill files after downloading?
Unzip the folder into ~/.claude/skills/ for a personal install, or into a project's .claude/skills/ folder for one project. You can also upload it in the Claude apps under Settings → Capabilities → Skills.
Why does the agent reply in Chinese?
The SKILL.md asks the agent to reply in Simplified Chinese unless the user explicitly requests English. Say in your request that you want English, and the agent should follow that.
Can I use this skill to recover passwords on my own computer?
The skill is not meant for that. Its purpose is sanctioned CTF and lab work, and general Windows administration or password recovery for your own accounts is better served by standard Windows tools or vendor documentation.
How does this compare to using the parent skill alone?
The parent ctf-sandbox-orchestrator sets sandbox assumptions and routes work. This skill is a narrower downstream workflow focused on DPAPI-protected secrets, masterkeys and the accepting service. Use the parent for general sandbox work and this one when the key secret is DPAPI-protected.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are a security student, competitor or lab trainer working on sanctioned Windows CTF exercises, and you already use the parent ctf-sandbox-orchestrator skill. Pick something else if you have no sandbox or authorisation, or if you only need standard Windows administration or recovery for your own accounts.
Download the skill