agentleFS
iktok90-designiktok90-design/ai-smart-contract-auditor/skills/access-control/SKILL.md

access-control

Skill36 starsChanged 5 months agoMITView on GitHub

Detect missing or incorrect access control — missing modifiers, wrong role checks, privileged function exposure, public initializers, and role-escalation paths. Activate on any function that mutates state, transfers funds, mints tokens, sets admin parameters, upgrades implementations, or pauses/unpauses.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use OwnableUpgradeable / AccessControlUpgradeable from OZ — never roll your own.

    line 81

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Access control detection

When this applies

Any state-mutating function. In particular:

  • Functions modifying balances, totalSupply, allowances, prices, fees
  • setOwner, transferOwnership, grantRole, setAdmin, setMinter
  • Upgrade pathways: upgradeTo, _authorizeUpgrade, proxy admins
  • Pause/unpause, emergency-withdraw, sweep, recoverERC20
  • Initializers (initialize, __Init, _init)
  • Functions guarded only by msg.sender == tx.origin or address checks against a single static value
  • Bridges, governors, vaults, anything with treasury

Detection patterns

Missing modifier (CRITICAL)

function mint(address to, uint256 amount) external {
    _mint(to, amount);   // ← anyone can mint
}

Wrong role check (HIGH)

function setFee(uint256 fee) external {
    require(msg.sender == owner || hasRole(USER, msg.sender));   // ← USER role can set fee
    fee_ = fee;
}

Initializer left public (CRITICAL)

function initialize(address admin) public {     // ← no initializer guard, anyone can re-init
    _grantRole(DEFAULT_ADMIN_ROLE, admin);
}

See also: [[initialization]]

tx.origin auth (HIGH)

require(tx.origin == owner);   // ← phishable via intermediate contract

Public privileged getter masking setter

Sometimes a setter is internal but a public wrapper exists with weak checks. Search for "alternate paths" to the same state slot.

Role admin self-grant (HIGH)

DEFAULT_ADMIN_ROLE can grant itself any role. If the admin is an EOA, a single key compromises everything. Look for renounceable admin patterns or multi-sig requirements.

selfdestruct reachable without ownership check (CRITICAL — but see [[selfdestruct-eip6780]])

Sweep / recoverERC20 with no asset allowlist (MEDIUM-HIGH)

function rescue(IERC20 token) external onlyOwner {
    token.transfer(owner, token.balanceOf(address(this)));  // ← sweeps any token incl. user deposits
}

Critical if it can sweep user deposits, Medium if only stuck airdrops.

Severity rubric

PatternSeverity
Privileged mint/burn with no authCritical
Public initialize on a deployed proxyCritical
Owner sweep that includes user depositsCritical
Wrong role guards a sensitive opHigh
tx.origin authHigh
Centralized single-key admin with no timelockHigh (see [[centralization-risk]])
Renounced ownership but admin role retainedMedium
Missing zero-address check on role grantLow

Remediation patterns

  • Use OwnableUpgradeable / AccessControlUpgradeable from OZ — never roll your own.
  • For each function: state the one role that should call it as a comment, then add the matching modifier.
  • Initializers: _disableInitializers() in the constructor, initializer modifier on init.
  • Multi-sig + timelock on DEFAULT_ADMIN_ROLE. Document the timelock duration.
  • Two-step ownership transfer (Ownable2Step) to prevent locking out the contract.
  • For sweeps: explicit allowlist of recoverable tokens, or denylist user-deposit tokens.

False-positive notes

  • Internal functions called only from properly-guarded externals are fine — verify call sites.
  • A function may look unguarded but its only side-effect is emitting an event — Info, not High.
  • "Permissionless" by design (e.g. claim() for caller) is fine — verify intent against the spec.

Related

  • [[initialization]] — initializer-specific access control
  • [[centralization-risk]] — admin-power audit
  • [[delegatecall-risks]] — proxy upgrade authorization

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.