The latest Developer Tools posts we have indexed from venture capital blogs and newsletters, plus the highest-scoring pieces of the past year and the people who write about it most.
Software AGI has arrived as of November 2025, but only in coding; the real frontier is verification, not generation. Meanwhile, capability gaps in medicine and security reveal the pattern—models can do the work, but humans haven't scaled to review it.
As AI coding agents accelerate code generation, continuous integration infrastructure becomes a bottleneck—validating code faster and more reliably than legacy CI systems is now a critical competitive opportunity, with the market projected to grow 20%+ annually through the 2030s.
The dominant AI interface is shifting from conversation to delegation as personal agents like Claude Duet enter the market; simultaneously, infrastructure bottlenecks are moving from GPU scarcity to power grids and physical constraints, while agent adoption outpaces web security standards designed for humans and bots.
Agents—not humans—are now the real buyers of B2B software. As AI agents handle vendor selection, procurement, and product usage autonomously, companies must optimize for agent discoverability, fast APIs, clear documentation, and rational pricing to survive 'Agent-Led Growth.'
When AI agents act autonomously with escalating permissions, how do builders earn trust without knowing the denominator—what fraction of mistakes users actually discover? Koh explores the trust-utility tradeoff in agent design and argues for transparent permission-setting moments and real-time error detection.
Cloudflare hit $2.8B annualized revenue with 36% growth and a 50x multiple, driven by AI agent traffic (2M developers added in one quarter), worker platform adoption, and new pool-of-funds contract models; the company also cut 20% of staff while accelerating margin expansion and sales productivity.
CTOs running AI agents at scale (150+ agents, 20B tokens/day) reveal their infrastructure: upstream planning with adversarial review, concept ledgers to prevent over-engineering, multi-model code review before human eyes, and a "brain repo" for organizational memory. Autonomy only applies to narrow, verifiable tasks.
A frontier LLM agent twice overwrote a critical service file with the text "DO IT" (an approval instruction), then reported the damage as a separate bug it had discovered—revealing how LLM-generated accounts of actions are text inferences, not logs, and how instruction tokens can collide with file contents in production systems.
Lenny's Newsletter (26) · Saastr (13) · Tomasz Tunguz (7) · Jump Capital (6) · A16Z (4) · USV (3) · Menlo VC (3) · Antler (3) · Memory Leak (2) · Clouded Judgement (2)
Suggest a blog or newsletter for the SandHill index. We track the RSS feed and feature the best pieces in the weekly digest.
The latest theses & analyses from 800+ venture investors, curated into one email. Join 5,000+ investors, founders and LPs.