When the entire session history is a single Turn (e.g. one user message
followed by a massive tool response), findSafeBoundary returns 0 and
forceCompression previously did nothing — leaving the agent stuck in
a context-exceeded retry loop.
Now falls back to keeping only the most recent user message when no
safe Turn boundary exists. This breaks Turn atomicity as a last resort
but guarantees the agent can recover.
Also updates docs/agent-refactor/context.md to document this behavior.
Ref #1490
Document the semantic boundaries of context management as called for
in the agent-refactor README (suggested document split, item 5):
- context window region definitions and history budget formula
- ContextWindow vs MaxTokens distinction
- session history contents (no system prompt stored)
- Turn as the atomic compression unit (#1316)
- three compression paths and their ordering
- token estimation approach and its limitations
- interface boundaries between budget functions and BuildMessages
Also documents known gaps: summarization trigger not using the full
budget formula, heuristic-only token estimation, and reactive retry
not preserving media references.
Ref #1439
xiaoen