query-degradation-is-deterministic-across-all-access-paths
OUT derived (depth 4)
All information access paths degrade gracefully while maintaining deterministic output: interactive queries cascade through tiered modes (LLM synthesis → bounded tool loop → raw FTS5 search), structured reads self-heal missing indexes via derived FTS5 reconstruction, and all fallback paths produce deterministic sorted output.
Summary
Every way of retrieving information from the system is designed to degrade through predictable fallback tiers rather than failing outright, and every fallback still produces sorted, reproducible output rather than nondeterministic results. This matters because it means the system can be trusted in CI pipelines and automated workflows — no matter what breaks (LLM unavailable, indexes missing), the output remains stable and useful.
Justifications
SL — Tiered query degradation and deterministic read resilience are independently established — together they show every information access path is both fault-tolerant and predictable.
Antecedents (all must be IN):
- all-query-operations-degrade-gracefully — All query operations degrade gracefully through multiple independent fallback tiers: ask cascades from LLM synthesis through bounded tool loops to raw FTS5 results on any failure, while search self-heals via index rebuilds on every save and falls back to substring matching on FTS5 unavailability — ensuring queries always return useful results regardless of LLM availability or index state.
- all-read-paths-are-deterministic-and-resilient — Every read-path operation produces deterministic output resilient to environmental variation: search handles missing FTS5 indexes via self-healing derived indexes and substring fallback, staleness checking produces sorted uniformly-structured CI-pipeline-ready results, and compact generates predictable bounded output from a pure function with fixed priority ordering.
Dependents
These beliefs depend on this one:
- query-resilience-serves-self-correcting-knowledge — All query access paths — interactive LLM synthesis, batch search, and compact summarization — degrade gracefully with deterministic output while operating against a knowledge base that actively self-corrects through contradiction resolution and staleness detection, ensuring degraded queries still return data from a consistency-maintained belief network