Live Daily intelligence for people building software in the AI era. It filters product news, technical change, and informed opinion down to what materially changes how builders ship.
Agentic AI · Editorial pipeline
Automated daily, reviewed weekly.
BuilderWithin and WPWithin are independent publications, designed and operated end to end on a multi-model agentic system. It selects candidates, verifies them against primary sources, drafts, reviews against editorial policy, and publishes on a schedule, unattended. Most of what it considers never ships.
2
Publications operated
5
Scheduled pipelines
Daily
Unattended releases
1 in 19
Candidates that ship
The publications
Live 
Live Practical WordPress intelligence for site owners, agencies, and hosting professionals. It turns important ecosystem, security, and product developments into clear guidance for people running sites day to day.
Selective by design
Generating articles was never the hard part. The hard part was building something that decides, reliably and without supervision, what should not exist. Most candidates are rejected, and the gates that reject them are the actual product.
Editorial pipeline
- 01IntakeCandidates arrive from parallel sources and are normalised into a single queue.
- 02SelectionAudience fit, editorial policy, and recent coverage decide what proceeds. Most candidates end here.
- 03Research and draftingA bounded agentic workflow verifies claims against primary sources and develops the piece, with distinct models routed to the work each suits.
- 04ValidationEditorial, factual, and structural checks run as hard gates. No path exists that publishes a failing draft.
- 05PublicationWork that clears every gate is released, and the run is recorded with its cost, decisions, and failures.
2,321considered
604passed selection
125published
5.4%publish rate
- Fallback lane
- Runs only when the primary filter is unavailable. Anything it recovers is independently validated before it rejoins the main path.
- Sample
- 84 automated BuilderWithin runs, 8 July to 6 September 2026. WPWithin runs the same design and rejects a comparable share.
How the system is built
Agent orchestration
Specialised agents with controlled handoffs across discovery, research, drafting, and review. Each is scoped to one job and cannot reach another agent’s tools.
Context engineering
Editorial policy, audience definition, recent coverage, and verified source material are assembled into the context behind every decision the system makes.
Multi-model routing
Distinct models are assigned to the work each suits, with a bounded fallback path that degrades safely instead of failing when a tier is unavailable.
Prompt and tool contracts
Role prompts, allowlisted tool permissions, and expected output shapes bound what any agent can do, read, or write.
Automated evaluation gates
Editorial, factual, and structural checks evaluate every draft before release. A failure discards the work rather than retrying it into acceptance.
Reliability and observability
Per-run cost, token, decision, and failure telemetry, alongside per-host circuit breakers, serialised release queues, and at-most-once delivery fences.
What the system will not do
The writer cannot publish
The drafting agent runs under an explicit permission contract with commit and release denied. Publication is a separate step it has no path to reach.
A failed check publishes nothing
Validation gates every run. Work that does not pass is discarded, never revised until it slips through.
No newsletter goes out unreviewed
The weekly issue opens as a draft for human review. Sending is a deliberate manual step, taken against the article already live in public.
Every run is inspectable
Each execution records its own cost, decisions, permission denials, and failures, so quality is audited rather than asserted.
Want to compare notes?
On the publications, the numbers, or what's next.
© 2026 James Lee · All rights reserved. · Privacy Policy