Search
Start AI Tools with Intent Solutions
The implementation journal of Intent Solutions. Follow the work of turning AI capability into reliable production systems: the builds, evaluations, failures, and operating lessons your team can use for itself.
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
Rank Current Truth with the Rerank, Not the Clock
Reindexing by updatedAt alone can let a chronological hit beat a current decision. The fix: lifecycle and history demotion in finalScore, with a paired test.
ReadGit Rebase Conflict: Verify Post-Push HEAD, Not Exit 0
On a shared checkout, the post-push HEAD, not the exit code, is what tomorrow's cron inherits. When a rebase conflicts mid-push, exit 0 hides a wedged tree.
ReadSieve :copy dies: rebuild the reply copy in the IMAP poller
Rebuild the inbox-reply copy loop in the IMAP poller with a named skip set as the safety net when the server-side Sieve redirect dies.
ReadRehearse the deploy, not the test suite
A production deploy rehearsal runs the real deploy scripts against the production Compose file. In catalyst-onboarding it caught a database data-directory permission bug the test suite could not.
ReadConnect every public surface to the new canonical
An evergreen earns the slot when home and About link to it. Search visitors land without knowing the company; the canonical must reach them on the first read.
ReadRequire the Phone at Intake, Not When Staff Call Back
now-lms /request-access added a required, server-validated phone field at intake (PR #123), stored as a labeled line of the existing body.
ReadTeam-Page Ordering Is a Release Decision
Team-page ordering is a release decision. Legal Counsel landed seventh; same-day hotfix; a 79-line CI gate keeps the roster authoritative across eight footers.
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 is an applied AI engineering company. We build with your team, test against your actual work, and transfer the knowledge needed to operate and evolve the system. This journal shares the methods and lessons that can be made public.
No commitment required. We respond within 24 hours.