JesseMarkowitzandClaude Opus 5
144406cd48
The release-validation milestone, and the thing it had to settle first was whether any of the earlier evidence meant what it said. M8 measured a deployment enforcing a 4,096-token input window while the application budgeted 16,384. Every request returned 200. What Ollama does with the excess is drop the oldest tokens, and the oldest tokens here are the system block — the narrator's rules and the campaign canon. A hundred-turn certification against that server would have looked perfect and proved nothing, which is why this milestone could not begin with a hundred turns. So the application asks now. Ollama's window is a property of how a model was loaded rather than of the request — sending num_ctx is accepted, ignored, and worse, reloads the model at the server's own default — so the only honest move is to find out and then tell the truth about it. /api/ps reports what a resident model is being served with, /api/show what an unloaded one will load with, both on the same host inference already uses, through the same endpoint policy and the same TLS trust store. A verified window is a ceiling on the budget; an unverified one leaves the budget alone and is recorded as unverified in the turn's own provenance, so an old turn can be asked afterwards whether it was built against a checked window. There is no third behaviour, and in particular no hard-coded 4,096: a number the server did not say would be right on one machine and wrong on the next. The proof that this is doing something is a campaign whose canon sits at the front of the prompt, 120 turns of history, and a 4,096-token window. The canon is still there afterwards and the oldest history is gone. The same campaign built the old way produces a prompt more than twice the window — the defect, reproduced, so the fix is measured against it rather than asserted. Two defects the validation found on its own, and they are the same defect twice: something was true and nobody was told. A manual state correction of four changes with one bad reference applied three, returned 201, and said nothing — while recording the refusal on the audit row nobody reads. It came to light because the identity diagnostic's own fixture was refused that way and the whole run proceeded on a campaign with no scene, which would have read as a model failure. And the narration-length setting moved no number: brief, medium and long each became one English sentence, while the numeric hint the model actually reads was derived from the global reply cap and said the same thing for all three. Both now say what they did. The other two post-M8 findings are closed as well. The tab said AI D&D, which no document had ever claimed it did not; it says Interactive Story now, with the open campaign first, and the name is the owner's decision rather than a find-and-replace to something narrower than the engine. After an Undo the reader could not tell where they had landed; the control row now ends with "Moment 11 · later story ahead", from the server's own answer, in the word the transcript already uses, with none of head, branch or depth anywhere near it. The identity diagnostic exists and the root cause does not. That campaign was destroyed, so no cause can be established — what M11 owes the finding is something that can classify the next occurrence, and a diagnostic that makes only the judgements a program can honestly make: duplicate keys, shared names, protagonist drift, state and context disagreeing. Whether prose misattributed a line is left to a person reading it beside its prompt, because a regex cannot read dialogue and one that pretended to would produce exactly the confident wrong answer this finding is about. Its detectors are proved to fire against a planted second Alice. Two entities may still share a display name. That was checked first, as the finding asked, and left permitted: a mother and a daughter, or a stranger giving a false name, are ordinary fiction, and refusing them to guard against a model mistake would refuse the wrong thing. What was missing was that it happened silently. It is reported now. Evidence, not inference: a hundred accepted turns against a real narrator with genuine process restarts; a real browser against the built SPA; a container with no network at all; a campaign moved into a data directory that never existed. Each was discarded and re-run whenever the product changed under it, and the runs that were thrown away are listed in the report with the reason, along with ten defects in the harnesses themselves — because a harness that has only ever agreed with itself is not evidence, and two of M8's five harness defects were masking real ones. No dependency was added, removed or upgraded. No acceptance test was retired, relaxed or reclassified. M11 is implemented and verified; it is not accepted, and there is no release tag. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Qyn3oRd4D6pi72nKBG725B
4.4 KiB
ChatGPT Project Sources — what to upload
This file exists for one purpose: to tell the repository owner which files
belong in the InteractiveStory ChatGPT project's Sources. It is not a
documentation index — planning/README.md is that.
planning/project-sources.txt is the same list as bare paths, one per line, for
gathering the files.
The rule behind the list: upload what is authoritative now. A project source is treated as current fact by whatever reads it, so a superseded document uploaded alongside a current one does not add context — it manufactures a conflict.
Refresh the Sources whenever the planning package is revised (VERSION.md
records each revision) and whenever a milestone completes.
REQUIRED PROJECT SOURCES
Twenty-eight files.
Repository orientation
README.md
DEVELOPMENT.md
PROVENANCE.md
The planning index and its version
planning/README.md
planning/VERSION.md
Product requirements and design
planning/SPECIFICATION.md
planning/TECHNICAL-DESIGN.md
planning/DATA-MODEL.md
planning/STORY-BRANCH-SEMANTICS.md
planning/CONTEXT-AND-MEMORY.md
planning/IMPORTED-KNOWLEDGE-DESIGN.md
planning/SECURITY-THREAT-MODEL.md
planning/MEDIA-EXTENSION-CONTRACT.md
planning/BROWSER-UX-SPEC.md
Implementation plan and acceptance contract
planning/BUILD-MILESTONES.md
planning/V1-ACCEPTANCE-TESTS.md
planning/TEST-CAMPAIGN-FIXTURE.md
The active ADRs — all eleven. They are short, and each one closes a question that will otherwise be reopened.
planning/DECISIONS/001-browser-first.md
planning/DECISIONS/002-ollama-only-v1.md
planning/DECISIONS/003-authoritative-local-state.md
planning/DECISIONS/004-local-only-production.md
planning/DECISIONS/005-branch-preserving-history.md
planning/DECISIONS/006-genre-agnostic-core.md
planning/DECISIONS/007-future-media-extension.md
planning/DECISIONS/009-ai-dnd-production-base.md
planning/DECISIONS/010-explicit-typed-narrative-state-events.md
planning/DECISIONS/011-local-inference-endpoint-policy.md
planning/DECISIONS/012-active-head-non-destructive-history.md
ADR 008 is deliberately absent: it was the process gate requiring Phase 0 to
close before production work began, Phase 0 closed on 2026-09-01, and it is now
in planning/archive/decisions/.
CURRENT-MILESTONE SOURCE
One file, and it changes as development progresses:
planning/reports/M11-IMPLEMENTATION-REPORT.md
M11 is the most recently completed milestone, and there is no next one to brief: what follows is independent review and the v1 acceptance decision. This report is the release-validation evidence package — its §F is the acceptance matrix, its §G the 100-turn campaign, its §O every defect found, and its §R the release-readiness answers. It is not an acceptance record.
Replace it, do not accumulate. When a later report lands, remove this one
from the project Sources and upload that one instead. The repository does the same thing:
planning/reports/ holds the current milestone's report and
planning/archive/milestone-reports/ holds the rest.
DO NOT UPLOAD AS PROJECT SOURCES
Not because these are worthless — because a source is read as current fact, and these are not current.
- Everything under
planning/archive/. The Phase 0 research reports, the Phase 0B recommendation and spikes, the completed M1 and M2 milestone reports, and archived ADR 008. Every conclusion they reached that still matters has already been applied to the active documents; what is left is superseded reasoning that will contradict the current package if uploaded beside it. - Older milestone reports, once their successor exists. Uploading M1, M2 and M3 together produces three descriptions of the same subsystem at three different stages.
- Source code.
backend/,frontend/, tests, migrations. The planning package describes the system; the code is read in the repository, where it can be searched and run. - Inherited upstream documentation. Upstream AI-DnD's
plan/build log anddocs/guides, generated HTML and screenshots were deleted from this repository on 2026-09-03 for exactly this reason: they describe a hosted, scripted, multi-user product this fork removed. Do not re-upload them from upstream or from Git history. - Generated HTML, images and screenshots. They add tokens, not facts.
repo-inventory.txtor any other local scratch output. Not tracked, not authoritative, stale the moment it is written.