Archprint is presented in the supplied case as a released tool that analyzes a TypeScript repository’s import graph, infers architectural boundaries, and turns those passing a statistical confidence gate into installable lint rules. The case attributes the creator’s claim to Tommkruix and frames the tool as a way to constrain coding agents without manually specifying architecture rules. The supplied web results cover adjacent dependency-linting and agent-instruction tools, but none directly documents Archprint; its release status, implementation, and effectiveness remain unverified here.
Archprint’s claimed inference-to-lint pipeline offers a concrete extension of Scott’s Semantic Decompilation and Test-First Agent Workflow: recover boundaries from existing code and turn them into executable checks rather than agent instructions. This merits testing whether inferred conventions deserve enforcement—not merely reflect existing mistakes—but the supplied material does not independently verify the tool or its effectiveness; the radar’s Vise episode tracks related invariant enforcement, not this inference mechanism.
ip:framework.semantic-decompilationip:concept.test-first-agent-workflowip:framework.architecture-not-vibesradar:vise-deterministic-refactor-gatesradar:concept.repository-intelligenceradar:concept.static-analysis
queries asked of Scott's wikis
- coding agent harnesses executable constraints versus prompt instructions
- repository architecture inference import graphs dependency boundaries
- automatically generated lint rules architecture drift
- statistical confidence gates inferred software conventions
- TypeScript projects architecture enforcement agent workflows
now 0 pts/hpeak 9 pts/hcomments 0/hpeers p31momentum: steady2 platformsage 642h
points/hour across evidence · reading as of 2026-10-05 04:37:03.724674+11:00 · deterministic, not a model opinion
2026-10-04T17:42:47Z
The new attachment is a third title-only, zero-comment RepoGuard post from the same single author under a second handle (now with a 'via MCP' framing) — the same duplicate-repost pattern already adjudicated as adjacent pattern interest, not corroboration of Archprint's inference-to-lint mechanism. After ~4 weeks of dormancy, trivial engagement across all threads, no independent implementation or evaluation, and the creator's own scoped no-drift finding bounding the agent-enforcement rationale, nothing new is within horizon and nothing is expected: the episode fades. The broader architecture-enforcement-for-AI-code pattern lives on elsewhere, but this case's specific claim is not developing.
2026-10-04T17:23:41Z
evidence attached: hn.story.49955746 — shared external link with case evidence
2026-10-03T19:42:02Z
The flagged attachment is a near-duplicate repost of RepoGuard — same underlying author under a second handle, title-only, zero comments — adding nothing since the prior look. The case's meaning is unchanged: an unvalidated inference-to-lint tool whose agent-enforcement rationale is already bounded by the creator's own scoped no-drift finding; it stays parked at watching/low despite the hot coding-agents neighbourhood, since none of that heat touches Archprint itself.
2026-10-03T19:26:18Z
evidence attached: hn.story.49946607 — shared external link with case evidence
2026-09-18T13:22:46Z
RepoGuard's title-only announcement indicates adjacent interest in architecture enforcement, not independent corroboration of Archprint's inference-to-lint mechanism or its value for coding agents. The attachment does not change the assessment: executable boundary recovery remains worth evaluating, but its correctness and workflow benefit are unvalidated.
2026-09-18T13:22:25Z
evidence attached: hn.story.49753346 — RepoGuard independently supports the developing pattern of enforcing repository architecture constraints around AI-generated code.
2026-09-12T20:29:03Z
The AgentRuleBench attachment names a relevant evaluation lead, but its title-only announcement from the same creator supplies neither results nor independent corroboration. It does not overturn the earlier scoped no-drift report or establish that Archprint's inferred constraints improve coding-agent workflows.
2026-09-12T20:21:51Z
evidence attached: hn.story.49676575 — AgentRuleBench directly tests whether agents violate architecture rules inferred from repositories, providing evaluation evidence for the existing architecture-enforcement episode.
2026-09-10T04:22:45Z
No substantive evidence has arrived to move Archprint beyond a creator-announced inference-to-lint tool; the echo is not independent corroboration. The scoped no-drift benchmark still limits its agent-specific rationale, while whether inferred boundaries merit enforcement remains an open evaluation question.
2026-09-08T03:25:11Z
The creator’s follow-up says the scoped benchmark found no agent architecture drift, weakening the motivating need for agent-specific enforcement without disproving the inference-to-lint mechanism. Archprint remains a concrete tool to evaluate for recovering executable boundaries, but neither the desirability of those boundaries nor its workflow benefit has independent validation.
2026-09-08T03:21:57Z
evidence attached: hn.story.49605147 — shared external link with case evidence
2026-09-08T00:26:21Z
No substantive new evidence changes the interpretation: Archprint remains a creator-announced inference-to-lint tool worth testing, not a validated architecture-enforcement workflow. The GitHub echo repeats the same claim rather than independently establishing implementation or whether inferred conventions deserve enforcement.
2026-09-08T00:25:34Z
grounded: converges/medium — Archprint’s claimed inference-to-lint pipeline offers a concrete extension of Scott’s Semantic Decompilation and Test-First Agent Workflow: recover boundaries f
2026-09-08T00:23:06Z
case created — A linked first-party tool provides a concrete architectural-enforcement mechanism distinct from existing repository-mapping cases, but adoption and effectiveness remain unestablished.