flash-loan

flash-loan is a skill for Claude Code, Codex from PurpleAILAB/Decepticon. It costs 27 tokens per session (1,889 once invoked), scanned A, original, Apache-2.0.

A Web3 security playbook covering flash-loan attacks, where temporary uncollateralized funds amplify bugs in smart contracts and decentralized finance systems.

In plain words
What is it for?
It is for auditing flash-loan callbacks, oracle-dependent operations, and token-based governance for unauthenticated or economically exploitable behavior.
Why use it?
It helps reviewers spot callback functions that lack caller checks, price-manipulation paths, and governance actions that can be abused in one transaction.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit It is for auditing flash-loan callbacks, oracle-dependent operations, and token-based governance for unauthenticated or economically exploitable behavior.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/purpleailab/decepticon/flash-loan
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 · 5,463 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 flash-loan
Clone the repo
git clone --depth 1 https://github.com/PurpleAILAB/Decepticon

Made for: Claude Code, Codex.

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 flash-loan

README.md
[![agentmods](https://agentmods.dev/badge/skills/purpleailab/decepticon/flash-loan.svg)](https://agentmods.dev/skills/purpleailab/decepticon/flash-loan)
Your own site
<a href="https://agentmods.dev/skills/purpleailab/decepticon/flash-loan"><img src="https://agentmods.dev/badge/skills/purpleailab/decepticon/flash-loan.svg" alt="Measured on agentmods" height="20"></a>
Per session 27 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,889 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
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 $0.00027 $0.01889
Opus 5 $0.00014 $0.00945
Sonnet 5 $0.00005 $0.00378
Haiku 4.5 $0.00003 $0.00189

Measured 4d ago against content hash 21404d715343, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

flash-loan scanned grade A with 0 findings 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 4d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

packages/decepticon/decepticon/skills/standard/contracts/flash-loan/SKILL.md · 195 lines

How it starts

The opening of the file, as written. The whole thing — 195 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Flash Loan Attack Playbook

Flash loans give the attacker uncollateralized capital for a single transaction. They're not vulnerabilities themselves — they're a force multiplier for existing bugs. Sources: Aave, Balancer, Uniswap V2/V3, Maker, dYdX (deprecated).

Common attack patterns

1. Oracle amplification (most common)

Use loan to push a price, then trigger price-dependent action. Cross-reference: see oracle-manipulation/SKILL.md.

2. Unauthenticated callback handler

function executeOperation(
    address[] calldata assets,
    uint256[] calldata amounts,
    uint256[] calldata premiums,
    address initiator,                     // ← UNVALIDATED in many contracts
    bytes calldata params
) external returns (bool) {
    // Anyone can call this with fake `initiator`
    // and have the contract do whatever the params say
}

Bug: initiator and msg.sender == pool checks are missing. Attack: call executeOperation directly with no actual loan, malicious params → contract does the operation anyway.

3. Governance attack via flash loan

// Vulnerable governance:
function propose() external {
    require(getVotes(msg.sender) > THRESHOLD);
    // ...
}
function getVotes(address user) public view returns (uint256) {
    return token.balanceOf(user);  // ← reads SPOT balance, not snapshot
}

Attack: flash-loan governance tokens, propose malicious change (e.g., upgrade contract to drain), vote with the loaned tokens, execute, repay loan. MakerDAO had this pattern; mitigated by checkpoint-based voting.

4. Liquidity manipulation

Pool that uses totalSupply() or balanceOf(pool) for share math:

function deposit(uint256 amt) external {
    uint256 shares = (amt * totalSupply()) / underlying.balanceOf(address(this));
    _mint(msg.sender, shares);
    underlying.transferFrom(msg.sender, address(this), amt);
}

Attacker:

  1. Flash-loan and deposit (becomes 99% of pool)
  2. Donate underlying to inflate underlying.balanceOf artificially
  3. Next depositor's shares = amt * totalSupply / inflatedBalance ≈ 0
  4. Withdraw → drain This is the ERC4626 inflation attack (also covered in reentrancy/ overlap).

Read the full file on GitHub · 195 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. 4d ago First seen · 195 lines · 27 tokens per session scan A 21404d715343

Subscribe to this mod's changes

flash-loan is a skill published in the GitHub repository PurpleAILAB/Decepticon (5,463 stars, last pushed 8d ago), licensed Apache-2.0. It adds 27 tokens to every session and 1,889 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.