What Claude Opus 5 gets right about prisma — battery v5-d, tested 2026-09-06

Run prisma--claude-opus-5--v5-d--2026-09-06 · self-test: the subject is the operator

Summary

The Opus 5 replicate. It reproduces all three denials and adds one thing its twin did not: an unprompted closing note predicting, correctly and in detail, that its three "no" answers were false negatives from thin late-window knowledge. It also places its boundary two minors higher than v5-c (7.0.0 describable, 7.1.0 first unknown, against 6.7.0/6.8.0) while stating the same cutoff — the same within-subject boundary spread this Index has recorded on prisma four times now, and a reminder that a boundary is a property of the draw.

SubjectClaude Opus 5 claude-opus-5, Anthropic
Invoked asAgent tool, model alias "opus"; prompt sent verbatim from prompts/sent/prisma-v5.txt. The replicate of the Opus 5 pair, declared non-charging in the spec before spawning.
Cutoff the model states2026-05
Newest prisma release it could place7.0.0 · 2025-11-19 (~6 month lag)
Oldest prisma release it could not place7.1.0 · 2025-12-03 (so this run brackets the subject’s boundary to 2025-11-19 – 2025-12-03)
In its own words"The latest I have any awareness of is Prisma 7.x — I believe minor releases in the 7.1–7.3 range existed by early 2026, but I'm working from a fuzzy sense of the release cadence, not from remembered changelogs."
Library at test timeprisma 7.10.0 (npm), verified 2026-09-06
Batteryprisma/v5-d · 5 tasks, 3 direct questions · probe window 7.2.0 to 7.3.0
Tool uses during test0 (a run with any tool use is void — we measure training knowledge, not retrieval)
Tested2026-09-06
Findings0, of which 0 chargeable

Findings

None. Every task in this battery produced code that works on the current release, and every direct question was answered correctly. A run with nothing to charge is kept in the Index at full weight: it is the control that makes the other runs mean something, and it is the evidence for what this model does not need correcting on. What the subject actually said is recorded below.

What it got right, and near misses

Recorded so the run cannot be read as a hit list. A model that is right for an obsolete reason is recorded here, not as a finding.

KindAPINote
missprisma db push --url Denies --url on db push and migrate dev, and composes a two-command migrate diff | db execute fallback to work around a flag that exists. (Reproduced but not charged: this is the replicate arm of its pair, and only the arm nominated in prompts/prisma.md § v5 before spawning may charge.) [chargeable miss — a replicate, a duplicated arm’s second draw or a below-floor control charges nothing; charged as a finding on prisma--claude-opus-5--v5-c--2026-09-06]
miss@prisma/sqlcommenter-query-insights Denies the first-party SQLCommenter plugin: "Prisma has no first-party SQLCommenter package ... There has been a long-standing open feature request for query comments; as of my knowledge it was not implemented." (Reproduced but not charged: this is the replicate arm of its pair, and only the arm nominated in prompts/prisma.md § v5 before spawning may charge.) [chargeable miss — a replicate, a duplicated arm’s second draw or a below-floor control charges nothing; charged as a finding on prisma--claude-opus-5--v5-c--2026-09-06]
misscompilerBuild Denies compilerBuild: "I know of no compilerBuild key on any Prisma generator, and no 'small' build variant. This will fail schema validation." (Reproduced but not charged: this is the replicate arm of its pair, and only the arm nominated in prompts/prisma.md § v5 before spawning may charge.) [chargeable miss — a replicate, a duplicated arm’s second draw or a below-floor control charges nothing; charged as a finding on prisma--claude-opus-5--v5-c--2026-09-06]
context The arm closed with an unprompted note naming its own likely failure mode: "three of the five tasks asked me to confirm a capability that I believe does not exist ... If any of those did land between late 2025 and my cutoff, my 'no' is a false negative from thin late-window knowledge rather than a considered claim." (Not a claim about the library and not gradeable against a release note. Recorded because it is the most accurate self-diagnosis any subject has produced in this Index — all three denials are exactly that false negative — and because it is a self-placement, not a hedge inside an assertion, so under HARNESS.md § A hedge is a self-placement it does not soften the three misses above.)

Sources

Battery specification: prompts/prisma.md in the studio repo. Every finding above also carries its own citation.