feat(recall): cap injected memory context (#71)

- Add `recall.maxCharsPerMemory` and `recall.maxTotalRecallChars` with defaults of `0`, which do not alter existing behavior. Users can opt in by setting positive values to cap injected memory context.
- Apply the budget after L1 search and before `<relevant-memories>` injection, preserving score order while truncating oversized entries and dropping overflow.
- Document the new guards in README, README_CN, and `openclaw.plugin.json`.

src/config.ts

@@ -80,6 +80,10 @@ export interface RecallConfig {
   enabled: boolean;
   /** Max results to return (default: 5) */
   maxResults: number;
+  /** Max characters injected for a single recalled L1 memory. 0 disables the per-memory limit. */
+  maxCharsPerMemory: number;
+  /** Max total characters injected for all recalled L1 memories. 0 disables the total limit. */
+  maxTotalRecallChars: number;
   /** Minimum score threshold (default: 0.3) */
   scoreThreshold: number;
   /** Search strategy (default: "hybrid") */
@@ -486,6 +490,8 @@ export function parseConfig(raw: Record<string, unknown> | undefined): MemoryTda
     recall: {
       enabled: bool(recallGroup, "enabled") ?? true,
       maxResults: num(recallGroup, "maxResults") ?? 5,
+      maxCharsPerMemory: num(recallGroup, "maxCharsPerMemory") ?? 0,
+      maxTotalRecallChars: num(recallGroup, "maxTotalRecallChars") ?? 0,
       scoreThreshold: num(recallGroup, "scoreThreshold") ?? 0.3,
       strategy: validateStrategy(str(recallGroup, "strategy")) ?? "hybrid",
       timeoutMs: num(recallGroup, "timeoutMs") ?? 5000,