Open Grant Proposal: AI Agent PKP Toolkit for Lit Protocol #89
Description
Name of Project:
AI Agent PKP Toolkit (Agent-first developer tools for Lit Protocol)
Proposal Category:
app-dev
Proposer:
aliveevie
Do you agree to open source all work you do on behalf of this grant and dual-license under MIT, APACHE2, or GPL licenses?:
Yes
Project Description
As AI agents become more capable and autonomous, a critical challenge emerges around secure identity, key management, and permissioned execution. Today, most AI agents either rely on insecure private key storage, centralized backend signing services, or overly permissive wallets. This creates significant security risks and limits the ability to deploy agents safely in production environments, especially for on-chain actions such as signing transactions, accessing encrypted data, or interacting with DAOs and protocols.
This project proposes building an AI Agent PKP Toolkit on top of Lit Protocol that enables developers to create AI agents backed by Programmable Key Pairs (PKPs) and Lit Actions, without ever exposing private keys. The toolkit will provide opinionated libraries and utilities that allow agents to authenticate, sign messages or transactions, encrypt/decrypt memory, and execute actions only when predefined conditions are met (e.g. DAO approvals, on-chain state, usage limits). By abstracting away low-level Lit primitives into agent-first developer tools, this project will significantly reduce integration complexity and improve safety.
The solution leverages Lit’s core strengths—programmable key management, access control, and decentralized execution—to deliver a secure, composable foundation for AI-native Web3 applications. Instead of building another agent framework, this project focuses on developer tooling and libraries that can be integrated into any existing agent stack, making Lit the default key and authorization layer for autonomous agents.
Value
This project creates strong strategic value for the Lit ecosystem by positioning Lit Protocol as the default security and key infrastructure for AI agents. As autonomous agents increasingly interact with blockchains, the need for programmable, non-custodial key management will grow rapidly. Getting this right enables Lit to capture a high-growth developer segment early and drive sustained PKP and Lit Actions usage.
If this problem is not addressed, developers will continue to deploy insecure solutions—hardcoded private keys, centralized signers, or overly permissive wallets—which increases systemic risk and slows adoption of Lit as a core infrastructure layer. Poor developer experience around agent integrations also risks Lit being bypassed in favor of simpler but less secure alternatives.
Execution risks include the evolving nature of AI agent frameworks and differences in developer workflows. This will be mitigated by keeping the toolkit framework-agnostic, focusing on well-documented primitives, and shipping clear examples and reference implementations that demonstrate correct and secure usage.
Deliverables
The final deliverable will be an open-source AI Agent PKP Toolkit consisting of:
- A core TypeScript/JavaScript library for integrating Lit PKPs into AI agent workflows
- Helper utilities for:
- PKP-backed agent identity
- Conditional signing using Lit Actions
- Secure encryption/decryption of agent memory and secrets
- Policy-based permissions (rate limits, on-chain conditions, DAO approvals)
- Pre-built Lit Action templates for common agent use cases
- Clear documentation and usage examples demonstrating:
- Agent signing flows
- Secure memory handling
- Conditional execution patterns
- Example reference implementations showcasing integration with popular agent architectures
Development Roadmap
Milestone 1: Architecture & Core Agent Identity Layer
Timeline: January 15, 2026 – February 5, 2026
Funding: $3,000
Team:
- 1 Lead Developer (ibxLab)
Scope & Deliverables:
- Define agent-first architecture for PKP usage
- Implement core library for PKP-backed agent identity
- Utilities for authentication and session management
- Initial documentation outlining design decisions and usage
Milestone 2: Conditional Signing & Lit Actions Toolkit
Timeline: February 6, 2026 – February 28, 2026
Funding: $4,000
Team:
- 1 Lead Developer (ibxLab)
Scope & Deliverables:
- Lit Action templates for conditional signing
- Support for on-chain state checks, DAO voting conditions, and usage limits
- Helper APIs to attach policies to agent actions
- Example flows demonstrating secure transaction signing by agents
Milestone 3: Secure Agent Memory & Reference Integrations
Timeline: March 1, 2026 – March 20, 2026
Funding: $3,000
Team:
- 1 Lead Developer (ibxLab)
Scope & Deliverables:
- Encryption/decryption helpers for agent memory and secrets
- End-to-end reference examples integrating the toolkit into an AI agent workflow
- Final documentation, README, and usage guides
- Open-source release and tagging of stable version
Total Budget Requested
Total Requested: $10,000 USD
Budget Breakdown:
- Core development & architecture: $6,000
- Lit Actions templates & security logic: $2,500
- Documentation, examples, and release support: $1,500
Maintenance and Upgrade Plans
ibxLab plans to maintain the toolkit beyond the grant period by incorporating it into ongoing internal R&D and client-facing infrastructure work. Future updates will track Lit Protocol upgrades, expand policy templates, and add support for emerging AI agent standards. Community contributions will be encouraged through clear contribution guidelines and issue tracking.
Team
Team Members
- Ibrahim Abdulkarim – Founder & Lead Developer, ibxLab
Team Member LinkedIn Profile
- Ibrahim Abdulkarim: https://www.linkedin.com/in/ibrahim193/
Team Website
Relevant Experience
ibxLab is a Web3-focused engineering studio building protocol-level tools, decentralized applications, and developer infrastructure. The team has experience delivering open-source tooling, on-chain automation systems, and secure smart contract integrations across multiple ecosystems. Ibrahim Abdulkarim has participated in and won multiple hackathons, built high-transaction on-chain applications, and specializes in designing developer-first abstractions for complex Web3 systems.
Team code repositories
Additional Information
-
How did you learn about the Lit Open Grants Program?
Through the Lit Protocol community and public grant documentation. -
Contact Email:
hello@ibxlab.com -
Additional Notes:
This project is designed to be framework-agnostic and composable, ensuring it can serve a wide range of AI agent builders while driving long-term adoption of Lit PKPs and Lit Actions.