Search
Start AI Tools with Intent Solutions
Daily work journal below. Real case studies, deep-dives, and operator-grade write-ups on AI engineering, DevOps, and the messy reality of shipping software that touches LLMs, by Jeremy Longshore at Intent Solutions.
Curated long reads
Let the Model Judge. Make the Code Decide.
The model owns judgment. Deterministic code owns the gates, the commit, and the audit trail. How I rebuilt a daily writing pipeline so a hung session can no longer brick tomorrow.
Adversarial Review: The Six Lenses That Halted a Rollout
Six lenses, live state, eighteen risks. I thought the safety work was shipped. The review proved three assumptions were theater. The durable rule: the model proposes; the system owns identity, trust, and state.
The Moat Is the Trust Layer
A local-RAG demo became Intent NEXUS. The moat is not the chatbot. It is egress control, code-enforced citations, a ledger you can check, injection defense, and evals that cannot lie quietly to you.
Recent posts
Auditing Written Claims Against Their Artifacts
A second review pass checks written claims against the artifacts they cite. It repaired about twenty statements of fact in records already treated as done.
ReadClaude Code Subagents Load at Session Start, Not at Commit
Claude Code registers project subagents at session start, so committed agents aren't available until restart. A human caught it, not the system.
ReadWhen a Gate Should Re-Run the Step Instead of Trusting Its Receipt
Have the gate re-run a deterministic step instead of trusting its receipt. Why the lander heals the record itself, and why uncommitted makes it safe.
ReadContract Tests Caught a 25 MiB Upload Truncation Reported as HTTP 201
Recorded-fixture contract tests found an upload truncated at 25 MiB and recorded as HTTP 201. The audit grade fell to B+ while coverage rose to 99.67%.
ReadPhoto to Listing: Barcode Decoding and LLM Re-rank
Exact issue identification is retrieval work, not vision work. How a photo to listing pipeline puts the LLM in the re-rank seat behind an evidence gate.
ReadEvery Claim Needs a Shipped Source and an Executable Proof
Every claim needs a shipped source and an executable proof. What a six-hour audit that closed nothing found out about code that was already finished.
ReadOne Corrected Check, Fifteen Repos
Correcting a gate is cheap. Propagating it through fifteen repos on a vendored, hash-pinned lane is the part that costs a day.
ReadSeries & curriculum
MCP for Beginners
End-to-end Model Context Protocol curriculum — solutions in Python, TypeScript, Java, Rust, C#, .NET — translated across six languages.
Agentic Design Patterns
Patterns and anti-patterns for production agent systems. Decision frameworks, prompt scaffolds, evaluation harnesses.
Tiny Recursive Models
Building small recursive systems where simple loops compose into emergent behavior. Series in progress.
Research & Curriculum
Long-form research articles, learning paths, and reading lists for the AI builder community.
Retrospectives
August 2026: A Control That Fires on Every Case and Changes No Outcome Is a Log Line
A control that fires on every case and changes no outcome is a log line, not a control. August 2026 retrospective: 31 of 31 posts flagged for inflation with the distribution unmoved, a learned cap rule that caught zero, and a rubric six weeks stale that still names a retired rule.
July 2026July 2026: Every Instrument Reported Success, and One of the Published Posts Was a 404
July 2026 retrospective: Tier 1 collapsed from 46% to 17%, the anti-inflation rule adopted July 3 fired zero times all month, the grader demoted half the catalog on a title keyword list, and a post with a full green audit trail was never published at all.
Let's build something that ships to production
Intent Solutions designs, builds, and ships production AI systems: Claude, MCP, RAG, agents, and the boring infrastructure that makes them trustworthy. Tired of AI demos that never make it past the POC?
No commitment required. We respond within 24 hours.