06-reference

founder bookmarks batch 1

Fri Apr 03 2026 20:00:00 GMT-0400 (Eastern Daylight Time) ·reference ·source: Founder bookmarks (curated reading queue) ·by Ben Wilson (founder)

Founder bookmarks — batch 1 (curated reading queue through Feb 2)

Why this is in the vault

The founder maintained a personal bookmark list of industry reading with short “why I saved this” notes for each link. Each item is either worth processing into its own vault entry or was captured as a time-stamped pulse on where the founder’s thinking was heading. Filed as a research queue so future sessions can pick items off one at a time and produce proper reference notes where the mapping against RDCO warrants.

Reading queue

Agent architecture

dbt-labs/dbt-agent-skills — dbt’s publicly-shared Claude Code agent skills. Founder note: “Good for my MG work. May adapt or compare to our existing agentic tooling.” → Candidate for own vault entry. Compare against RDCO’s 22-skill layout and the /audit-model / /generate-tests skills. Relevant to 2026-04-11-garry-tan-thin-harness-fat-skills.

@trq212 on how Anthropic thinks about skills internally — tweet about Anthropic’s internal skill discipline. Founder note: captured for industry pulse. → Candidate for own vault entry via xmcp. Short but canonical source.

Snap Engineering: agent format — Snap’s structure for agents. Founder note: “not sure how much of it will pertain to us here, but capturing as part of the developing industry thoughts.” → Medium priority. Industry pulse on agent-format conventions.

@jolandgraf on remote background agents — Founder note: “im definitely thinking about this. How to get agents off of my machine, but still accessible. Claude Code channels has been a big step in that direction for me… I haven’t fully bought into remote background agents - but maybe this article can convince me.” → Candidate for own vault entry. Directly related to the ../04-tooling/rdco-state-ownership-architecture question of where state lives. Also connects to 2026-04-13-jaya-gupta-ai-lock-in-state-moat.

Data engineering

rmoff: Claude Code isn’t going to replace data engineers yet — Founder note: “lots of people are concerned about ai replacing white collar work. Data engineering fits into that and is my profession, so wanted to understand the arguments made here. I’m thinking lean into this technology. Augmentation, not replacement, but I’m not sure if that aligns with the argument the article is making.” → High priority for own vault entry. Directly relevant to the RDCO positioning and the phData vs MG decision. The augmentation-vs-replacement framing is load-bearing for the agent-deployer consulting posture (see 2026-04-14-levie-agent-deployer-role-jd).

dedp.online — Deep Data Engineering Playbook — Freely available book. Founder note: “We want to come to our own conclusions, but this should be excellent material to feed into building up our knowledge base.” → Processing candidate via /process-book skill (when built). Will produce multi-chapter synthesis like the ADWD and Halls-Moore books.

Medium: Ontology, Taxonomy, Data Model, Context Graph & Friends — Founder note: “some thoughts about how to structure knowledge graphs.” → Relevant to the ../01-projects/graph-db-eval/vertex-edge-dictionary work. Worth processing.

Marketing / content strategy

Every: Compound Engineering — Founder note: “Something I think we are already doing, but good to document the source of the term and see how our setup compares.” → Candidate for own vault entry. Puts a name on what RDCO is doing. Links to content series planning.

@startupideaspod on marketing ideas — Founder note: “marketing is something I’ll definitely need your help with! This isn’t a definitive article, but something we can start thinking about.” → Low priority; placeholder flag for the marketing workstream.

@vibemarketer_ on using Claude for marketing — Founder note: “another article on how to use Claude for marketing.” → Low priority; pulse-check on how others are using Claude in the marketing workflow.

Processing notes