Ground newsroom claims in asserted public facts
This commit is contained in:
parent
a585398319
commit
02fcc7380f
9 changed files with 245 additions and 68 deletions
|
|
@ -486,31 +486,30 @@ world whose derived cover grew from four cells to six as named figures emerged;
|
|||
the 200-cell path remains a local scale proof and the
|
||||
Delvehold game-side adapter is not implemented here.
|
||||
|
||||
An isolated two-wave connector run against Ghostlight source `2f7c090` advanced
|
||||
one fresh campaign from revision 0 to 2, committed 19 events and 14 news rows,
|
||||
and left player state and location unchanged. Its reader pages and separate
|
||||
provenance artifacts live at
|
||||
`F:\GameCult\GhostlightDungeon\acceptance\root-wire-frontpage-blind-2f7c090-1`.
|
||||
The wave-two page did not pass the blind journalistic gate. A fresh reviewer,
|
||||
shown only that page under a neutral brief, diagnosed procedural construction
|
||||
with 99 percent confidence because the copy exposed a canonical relation name,
|
||||
a numerical state delta, and source-record language. This does not weaken the
|
||||
world-simulation witness. It establishes that rendering and copy-desk gates
|
||||
cannot repair a newsroom packet that presents mechanical background records as
|
||||
editorial evidence.
|
||||
The first newsroom projection cut was committed as `a585398` and exercised by
|
||||
a fresh isolated two-wave Codex Connector run with `gpt-5.6-luna`. The campaign
|
||||
advanced to revision 2 with 18 Events and 15 NewsIssues while player state and
|
||||
location remained unchanged. Both presentation gates failed: the blind editor
|
||||
returned FAIL at 99 percent confidence and the grounding auditor returned FAIL
|
||||
at 0.94 confidence. The untouched responses and their review hashes live at
|
||||
`F:\GameCult\GhostlightDungeon\acceptance\root-wire-frontpage-blind-a585398-1`.
|
||||
Public Event prose still exposed abstract posture and relation transitions; the
|
||||
newsroom promoted merely involved entity and location metadata into asserted
|
||||
names and a dateline; and the smoke result projected the wave-one commit at top
|
||||
level beside wave-two counts and head. The world run remains valid evidence of
|
||||
autonomous transition, not newspaper acceptance.
|
||||
|
||||
The current local source cut rewrites public strategic summaries into
|
||||
human-readable accounts at the event owner, gives the editor only opaque desk
|
||||
citations and named public facts, deduplicates identical fact packets, and maps
|
||||
accepted citations back to exact canonical `NewsIssue` and `Event` provenance.
|
||||
All 392 library tests and every binary test pass, and the bounded Modeling
|
||||
re-audit is clear. This is local verification only; no fresh provider page has
|
||||
passed blind or grounding review yet.
|
||||
A second local cut changes institution and relation public phrasing, includes
|
||||
model-facing names only when the public account asserts them, suppresses
|
||||
ambiguous multi-place metadata, and projects final-wave commit data at the smoke
|
||||
result top level. All 393 library tests and every binary test pass, and the
|
||||
bounded Modeling final review is clear after correcting a name-boundary
|
||||
collision. The cut is not yet committed or run against the provider.
|
||||
|
||||
1. Commit and push the coherent newsroom and public-event cut.
|
||||
1. Commit and push the exact locally verified second cut.
|
||||
2. Run a fresh isolated exact-commit two-wave newspaper acceptance through
|
||||
Codex Connector with `gpt-5.6-luna`, preserving world, issue, audit, receipt,
|
||||
build, and hash witnesses without mutating the live daemon.
|
||||
build, review, and hash witnesses without mutating the live daemon.
|
||||
3. Dispatch a fresh blind reviewer with only the strongest page and a neutral
|
||||
editorial brief; preserve its untouched response.
|
||||
4. Independently audit that page against committed public events before
|
||||
|
|
|
|||
|
|
@ -939,7 +939,11 @@ machinery never become the public account. The newsroom does not re-decide
|
|||
which committed public events count as truth. It merges identical fact bundles
|
||||
under one opaque per-request citation while retaining every backing news row,
|
||||
event, channel, and reliability record for later audit. The editor receives
|
||||
only those citations, human-readable accounts, and canonical names. News IDs,
|
||||
only those citations, human-readable accounts, and canonical names asserted by
|
||||
those accounts. Event membership metadata does not silently turn an involved
|
||||
actor, institution, population, or ambiguous multi-place scope into an asserted
|
||||
fact or dateline; a single event location may supply its canonical dateline,
|
||||
while a multi-place event must name the place in its public account. News IDs,
|
||||
event IDs and kinds, channels, reliability, source headlines, and wall-clock
|
||||
timestamps remain internal. A capable editor model owns selection,
|
||||
grouping, headline, deck, byline class, dateline, and article prose as a
|
||||
|
|
@ -963,7 +967,10 @@ strategic acceptance smoke use that same primitive before returning a
|
|||
rejection. In the acceptance smoke, an interim edition rejection is local to
|
||||
the consumer and cannot stop later world waves. A rejected final combined
|
||||
edition remains a nonzero acceptance failure, but first writes the completed
|
||||
wave reports, rejection receipts, and failed status. The reader renderer has no
|
||||
wave reports, rejection receipts, and failed status. Its top-level plan,
|
||||
receipt hash, and commit always project the final completed wave, matching the
|
||||
top-level counts and persisted campaign head; per-wave history remains under
|
||||
`waves`. The reader renderer has no
|
||||
path to provenance fields and escapes all model- and consumer-supplied plain
|
||||
text before Markdown emission; the provenance renderer applies the same
|
||||
plain-text boundary to its audit data. Neither editor, copy desk, renderer, nor
|
||||
|
|
|
|||
Loading…
Reference in a new issue