Does compiled opponent state retain value across seeds, held-out policies, Context loss, and Harness replacement?
Distinguish durable strategic information from fixture-specific explanation and provider-specific reasoning style.
01 / Current position
A hypothesis is not the current judgment.
The dossier preserves the live research position. Dated articles preserve the complete evidence and argument that changed it.
Explicit opponent hypotheses and revisions will improve held-out switch recognition, resource timing, and objective completion when transcript Context is truncated or the Harness is replaced, provided the state is compiled from verified observations rather than model narrative alone.
Open after mixed Round 1 evidence. The local opponent-aware actor improved diagnosis and objective rate, one Codex strategic Trial recognized a switch, Hermes outcomes were unchanged, and no model Trial completed the genuine objective. No causal or transferable advantage is established.
02 / Decision boundary
What keeps this Question alive?
Round 1 improved local switch recognition and preserved richer hypotheses, but all retained model Trials failed the objective. Without transfer, opponent state is descriptive overhead rather than durable capability.
Compare transcript-only and compiled-state conditions across multiple seeds, held-out opponent policies, bounded Context loss, and both Codex and Hermes Harness replacement.
A reactive policy or transcript-only Agent matches held-out objective, information, timing, and recovery outcomes with lower state and maintenance cost.
03 / Supporting publications
Complete arguments connected to this Question.
1 dated publication currently document this research line.
04 / Source discipline
The dossier is an index, not the evidence authority.
Owns the current judgment, next test, and deletion condition.
Own complete dated arguments, limitations, comparisons, and source links.
Own exact code, tests, releases, receipts, and machine evidence.