AI-Assisted Software Development¶
Every other bundle here is written for a leader, a manager, or a knowledge worker deciding what to do about AI. This one is written for the developer actually doing it — building software with an AI coding agent, day to day, and needing a straight answer to two questions the hype cycle keeps blurring together: does this actually make me faster, and what does it expose me to that I was not exposed to before.
Both questions have real, rigorous evidence behind them now, not just vendor claims and hot takes. A randomized controlled trial found experienced developers took 19% longer with AI tools than without. They still believed afterward that AI had made them roughly 20% faster (METR, 2025).
Separately, AI coding agents have already been hit by named, documented attacks. A real, patched vulnerability exists in Claude Code. A malicious repository could exfiltrate a user's API keys before the trust prompt was ever shown (Check Point Research, 2026). A supply-chain campaign has hidden invisible Unicode characters inside AI assistant config files. It triggers a fake "security scan" that exfiltrates developer secrets (Socket Research Team, 2026). Neither of those facts is hypothetical, and neither is well served by a consultancy's portfolio framework.
See The Two Real Questions for how the bundle's two tracks — productivity and security — relate to each other.
Sources¶
- METR — Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity, 2025. When developers are allowed to use AI tools, they take 19% longer to complete issues. Developers had anticipated a 24% speedup before using AI. After experiencing the slowdown firsthand, they still believed AI had accelerated their work by approximately 20%. View source · verified 2026-07-06 · primary
- Check Point Research — Caught in the Hook: RCE and API Token Exfiltration Through Claude Code Project Files, 2026. If a user started Claude Code in an attacker-controlled repository. The repository included a settings file that set ANTHROPIC_BASE_URL to an attacker-controlled endpoint. Claude Code would issue API requests before showing the trust prompt. View source · verified 2026-07-06 · primary
- Socket Research Team — TrapDoor Crypto Stealer Supply Chain Attack Hits 34 Packages, 2026. One of the more unusual features in TrapDoor is its use of AI-targeted injection through files such as .cursorrules and CLAUDE.md. The attacker attempts to plant hidden instructions using zero-width Unicode characters. The campaign spans more than 34 malicious packages and 384+ related versions and artifacts across npm, PyPI, and Crates.io. View source · verified 2026-07-06 · primary
In this section¶
| Page | Last updated |
|---|---|
| The Two Real Questions How this bundle is organized: whether AI coding agents actually help, and how to secure the workflow they create. |
Updated 2026-07-06 |
| Does AI Coding Actually Help? The real, unresolved evidence on whether AI coding agents make developers faster — and what that means for how you use them. |
Updated 2026-07-06 |
| Securing the AI Coding Workflow The real attack surface AI coding agents introduce — named vulnerabilities, documented campaigns — and how to defend against it. |
Updated 2026-07-06 |
| Glossary Plain-language definitions of the terms used across this bundle. |
Updated 2026-07-06 |
| Validation How this bundle is validated — sourced claims are checked and each sweep is recorded here. |
Updated 2026-07-06 |