Add fast cache-first narration mode for play runs.
Shipped as ad4ec82.
What Changed
This entry documents NPC narration work, including cache behavior and controls that keep AI-backed flavor usable in play.
Shipped ad4ec82.
What changed
- Narrated mode now defaults to fast mode:
- cache hits are used immediately
- max 1 uncached live Letta call per run pass
- remaining uncached beats are deferred instead of stalling the page
- Full narration is still available by turning fast mode off.
Why
Under slow model responses, full live narration across multiple beats could push the request into gateway timeout territory. This keeps runs responsive while preserving cached dialogue continuity.
Verification
php tests/run-tests.php-> 53 passed- local smoke: fast-mode UI + run result renders
- prod smoke: narrated endpoint returns normally in fast mode
Why It Matters
AI narration needs operational discipline. The player should get richer texture without waiting on slow live calls or losing deterministic replay behavior.
Implementation Notes
The commit message for this slice was:
Narrated runs now default to a one-call live Letta budget with cache reuse across remaining beats, reducing timeout risk while keeping optional full narration available.Files touched in this slice included:
public/includes/Letta/NpcIntentBroker.phppublic/play.php
Verification
- The entry is part of the devlog archive and renders through the same PHP Markdown pipeline as current posts.
- The test suite now requires every devlog post to carry substantive details, illustration and screenshot figures, and valid visual asset references.
- Current browser smoke checks cover representative rich posts and older auto-generated posts after expansion.
Visuals
