account-abstraction

account-abstraction is a skill for Claude Code, Codex from keep-starknet-strange/starknet-agentic. It costs 28 tokens per session (673 once invoked), scanned A, original, MIT.

Guidance for account abstraction in Starknet, where smart contracts can act as user accounts and handle transaction checks. It focuses on validation, execution, nonces, signatures, replay protection, and session-key policies.

In plain words
What is it for?
Use it to review or design account validation and execution paths, session-key rules, nonce handling, signature checks, and privileged selector protections.
Why use it?
It helps prevent account contracts from accepting unauthorized actions, replayed signatures, unsafe calls, or costly validation requests.

Skill for Claude CodeCodex

Written for Claude Code and Codex: allowed-tools in frontmatter, but also agents/openai.yaml present.

Part of the starknet-agentic-skills plugin — 19 skills, 2 commands, 1 plugin shipped together

Good fit Use it to review or design account validation and execution paths, session-key rules, nonce handling, signature checks, and privileged selector protections.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/keep-starknet-strange/starknet-agentic/account-abstraction
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 keep-starknet-strange/starknet-agentic --skill account-abstraction
Clone the repo
git clone --depth 1 https://github.com/keep-starknet-strange/starknet-agentic

Made for: Claude Code, Codex.

Or install starknet-agentic-skills, the plugin that ships this one along with the rest of its 19 skills, 2 commands, 1 plugin.

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 account-abstraction

README.md
[![agentmods](https://agentmods.dev/badge/skills/keep-starknet-strange/starknet-agentic/account-abstraction/github.svg)](https://agentmods.dev/skills/keep-starknet-strange/starknet-agentic/account-abstraction)
Your own site
<a href="https://agentmods.dev/skills/keep-starknet-strange/starknet-agentic/account-abstraction"><img src="https://agentmods.dev/badge/skills/keep-starknet-strange/starknet-agentic/account-abstraction/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 account-abstraction

Your own site · 80×15
<a href="https://agentmods.dev/skills/keep-starknet-strange/starknet-agentic/account-abstraction"><img src="https://agentmods.dev/badge/skills/keep-starknet-strange/starknet-agentic/account-abstraction.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 28 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 673 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 $0.00028 $0.00673
Opus 5 $0.00014 $0.00336
Sonnet 5 $0.00006 $0.00135
Haiku 4.5 $0.00003 $0.00067

Measured 11d ago against content hash c9aac51051cc, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

account-abstraction 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 11d 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.

skills/account-abstraction/SKILL.md · 76 lines

How it starts

The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Account Abstraction

When to Use

  • Reviewing account contract validation and execution paths.
  • Designing session-key policy boundaries.
  • Validating nonce and signature semantics.

When NOT to Use

  • General contract authoring not involving account semantics.

Quick Start

  1. Confirm __validate__ enforces lightweight, bounded checks.
  2. Confirm __execute__ enforces policy and selector boundaries.
  3. Verify replay protections (nonce/domain separation) for all signature paths.
  4. Add regression tests for each fixed session-key or policy finding.
  5. Run cairo-auditor for final AA/security pass before merge.

Core Focus

  • __validate__ constraints and DoS resistance.
  • __execute__ policy enforcement correctness.
  • Replay protection and domain separation.
  • Privileged selector and self-call protection.

Workflow

References

starknet.js Example

import { Account, CallData, RpcProvider } from "starknet";

const provider = new RpcProvider({ nodeUrl: process.env.STARKNET_RPC! });
const account = new Account(provider, process.env.ACCOUNT_ADDRESS!, process.env.PRIVATE_KEY!);

// Validate preview (debug-only): inspect __validate__ behavior with the current nonce.
const nonce = await account.getNonce();
const call = { contractAddress: process.env.TARGET!, entrypoint: "set_limit", calldata: CallData.compile({ value: 7 }) };
await provider.callContract({
  contractAddress: account.address,
  entrypoint: "__validate__",
  calldata: CallData.compile({ calls: [call], nonce }),
});

// Execute path: real transaction that triggers __execute__ and nonce checks.
const tx = await account.execute([call]);
await provider.waitForTransaction(tx.transaction_hash);

Error Codes and Recovery

Code Condition Recovery
AA-001 __validate__ is too expensive or stateful Remove heavy logic from validation; add a test that caps validation steps.
AA-002 __execute__ allows blocked selectors/self-calls Enforce selector filters and self-call checks; add authorized/unauthorized regression tests.
AA-003 Nonce or domain mismatch causes replay risk Normalize nonce source/hash domain; add replay and cross-domain tests.
AA-999 Unexpected runtime panic Capture calldata + caller context, reproduce in unit tests, then escalate to cairo-auditor.

Read the full file on GitHub · 76 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 11d ago First seen · 76 lines · 28 tokens per session scan A c9aac51051cc

Subscribe to this mod's changes

account-abstraction is a skill published in the GitHub repository keep-starknet-strange/starknet-agentic (80 stars, last pushed 3d ago), licensed MIT. It adds 28 tokens to every session and 673 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 2026-08-30.