← Kingy Local AIv23 stopped after 44 of 240 planned originals
document-comparison-v23-20261018.1 · candidate 1.5.6
Incomplete and not graded. Candidate 1.5.6 remains held; the active measured workflow remains 1.4.0.No new answer-quality result, head-to-head advantage, latency comparison or overall winner is established.
What happened
Kingy's input-only helper returned after one input call while preparing native B1 original5. Its subsequent draft-persistence observation failed during JSON parsing. The preserved error hash matches SyntaxError: Unexpected end of JSON input.
The same prepared draft was later observed intact, with zero native messages, zero Send calls and no dispatch journal for original5. Preparation of that draft does not count as a fifth completed native original or a fifth answer-generation request. The study stopped without retrying the question through a different route.
The underlying cause is unknown. Failed read bytes and the stack were not retained. A transient incomplete backing-file write is an inference only, not a verified diagnosis. This Kingy observation/control failure does not establish LM Studio product unreliability or inferior answer quality.
Retained operational accounting
Operational metadata reviewed by a separate AI agent| Planned originals | 240 |
|---|
| Completed Kingy originals (A1) | 40 |
|---|
| Completed native LM Studio originals | 4 |
|---|
| Completed originals, total | 44 |
|---|
| Unexecuted planned originals | 196 |
|---|
| A1 execution errors / timeouts | 0 / 0 |
|---|
| A1 answer-generation requests | 90 |
|---|
| A1 document-free warmup requests | 1 |
|---|
| B1 model loads / document-free warmups | 1 / 1 |
|---|
| B1 total input calls / Send calls | 5 / 4 |
|---|
| Fifth original: prepared input-only drafts | 1 |
|---|
| Fifth original: native messages / Send calls | 0 / 0 |
|---|
| Fifth original: dispatch journal exists | No |
|---|
The 40 Kingy originals completed without recorded execution errors or timeouts. The four native originals have accepted first-user and backend-completion proofs. None of these completion counts establishes answer correctness. The 90 Kingy answer-generation requests belong to its 40 originals and are not 90 completed originals. All frozen scientific source files remained unchanged.
What remains unrun
- The full 240-original comparison and its answer-quality grading were not completed. Reference values remained closed.
- Numeric quality qualification, scorer evaluation, fresh package selftests, package authentication and denied-egress qualification checks were not run for this study. These are unrun checks, not failed answer-quality or security tests.
- No general accuracy, speed, product reliability, privacy, cost, hardware-upgrade or competitor-superiority claim follows.
Evidence and preservation
Actual root closure, owned generation-model/logger cleanup and a 291-check operational review are bound by receipt hashes. A separate AI agent checked operational metadata on this same Mac. Human validation, reproduction on another machine and semantic answer-quality validation were not performed. Any later repair is prospective and cannot establish an outcome for this retained study.
Original evidence remains preserved locally. These four public files contain the exact approved safe operational projection and hash lineage. They exclude raw answers, question/reference values, grading keys, native backings, private UI, source code and private paths.
Exact approved safe operational summary · Receipt and preserved-study hash lineage · Published file hashes
Preserved v19 study · Preserved v20 study · Preserved v21 study · Preserved v22 study. Their evidence and limitations remain separate.