Affaan Mustafa
Affaan Mustafa is an individual, solo developer and creator who builds at the intersection of AI, finance and developer tools. His work focuses on agentic systems, open-source coding-agent infrastructure and compute markets.
At a Glance
- Individual developers and open-source maintainers using AI coding agents
- Software engineering teams and enterprises using Claude Code, Cursor, Codex, OpenCode or related tools
- Security-conscious teams auditing agent configuration and software supply-chain risk
- GPU buyers, sellers and infrastructure operators
- +2 more
AI Tools by Affaan Mustafa
(2)
AgentShield
AI Agent Security Scanner
ECC Tools (Everything Claude Code)
Agent Harness Engineering Toolkit
Discussions
No discussions yet
Be the first to start a discussion about Affaan Mustafa
Latest News
Everything Claude Code repository continued active development, with the latest commit recorded on September 24, 2026.
AgentShield released v1.6.0, with the repository describing 268 rules across 15 modules and a rebuilt distribution.
Tech Times reported that ECC had reached more than 228,000 GitHub stars and described its expansion to 67 agents, 278 skills and seven AI coding tools.
ECC reached the v2.0.0 cross-harness milestone, extending native support across Claude Code, Codex CLI, Cursor, OpenCode, Gemini, Zed and GitHub Copilot.
Products & Services
Open-source agent harness/configuration system containing skills, specialized agents, hooks, rules, commands, memory persistence and workflow patterns for coding agents. The repository supports Claude Code, Cursor, Codex, OpenCode and other harnesses; a hosted GitHub App/ECC Pro offering supports private repositories.
Open-source security auditor for AI-agent configurations. It scans Claude Code and related agent configuration for hardcoded secrets, unsafe permissions, hook injection, MCP-server risks and prompt-injection vectors. Distributed as the ecc-agentshield npm CLI, GitHub Action and GitHub App integration, with MiniClaw APIs.
Agentic desk/marketplace for GPU compute. It prices and arranges buying, selling or hedging of GPU capacity against published indices and forward curves, with fixed-price term agreements and an agent-facing compute CLI.
AI-powered customer-discovery platform created with David Rodriguez. It researches an ideal customer profile, generates synthetic customer personas, enables conversations with those personas and supports real-customer validation before product development.
Market Position
Affaan's work is positioned as infrastructure around AI agents rather than another general-purpose model. ECC differentiates from ad-hoc prompts and separate per-tool configurations by providing a cross-harness, open-source system of agents, skills, hooks, memory and rules; its alternatives include native Claude Code/Cursor/Codex configurations and other agent-harness/toolkit projects. AgentShield differentiates by focusing on the security of agent configuration files and workflows, competing with general code-security scanners and emerging AI-agent security tools. Itô positions itself as a market/OTC desk and risk-transfer layer for GPU compute rather than a conventional cloud GPU rental marketplace.
Leadership
Founders
Affaan Mustafa
Builder and creator; founder of DCUBE, creator of Everything Claude Code, AgentShield, Zenith Chat and several autonomous trading/developer-tool projects. He studied Mathematics-Computer Science and Business Economics at UC San Diego, later attended the University of Washington Applied/Computational Mathematics graduate track, and holds an MIT Applied Data Science Professional Certificate.
Executive Team
Affaan Mustafa
Co-founder of Itô; creator of Everything Claude Code and AgentShield
SF-based builder working across AI, markets and research; previously founder of DCUBE, head of product and founding engineer at PMX Trade, core developer at elizaOS, and creator of Zenith Chat and other agentic trading projects.
Founding Story
Affaan describes himself as a builder who started shipping products at 18. He founded DCUBE to use customized transformer models to remove repetitive CRM and administrative work and bootstrap tuition, then moved through AI-agent, trading and developer-tool projects. His current direction is Itô, an agentic desk for compute, alongside open-source agent infrastructure such as Everything Claude Code and AgentShield.
Business Model
Revenue Model
Open-source projects are distributed under MIT-style open-source terms and supported by sponsorships. ECC also has a hosted GitHub App/ECC Pro paid tier for private repositories and consulting for enterprise rollouts, agent-harness performance and security reviews. Itô operates as a compute desk that arranges and matches fixed-price GPU compute trades; its agentic compute CLI is intended to let users find, reserve and run compute from inside an agent.
Pricing Tiers
Hosted GitHub App with private-repository support, as described on Affaan Mustafa's GitHub profile and in Tech Times.
Target Markets
- Individual developers and open-source maintainers using AI coding agents
- Software engineering teams and enterprises using Claude Code, Cursor, Codex, OpenCode or related tools
- Security-conscious teams auditing agent configuration and software supply-chain risk
- GPU buyers, sellers and infrastructure operators
- Founders and product teams conducting customer discovery
- Real-estate, insurance and other SMB operators needing CRM/admin automation
- AI-assisted software development and engineering workflow standardization
- Test-driven development, code review, security review and build-error resolution with coding agents
- Auditing Claude Code and other AI-agent configurations before deployment
- Enterprise/private-repository agent-harness rollouts and consulting
- Buying, selling and hedging GPU compute capacity
- Customer discovery, product-market-fit research and synthetic user interviews
- Zenith Chat's target users are founders seeking pre-build customer and product-market-fit validation.
- DCUBE targets small-to-medium businesses, especially in real estate and insurance, that lack internal optimization or data teams.