fix(agent): detect repeated tool call loops and make max_iterations configurable

Local LLMs often get stuck calling the same tool repeatedly because they
lack the sophistication to synthesize results. The agent loop had no
safeguard — it re-executed whatever the model requested up to 10 times.

Add fingerprint-based loop detection: if the same tool+args combination
repeats 3 consecutive times, break the loop and return the last results.
Also add agents.max_iterations to the config schema so the iteration
limit is user-configurable (default: 10).

src/config/schema.ts

@@ -220,6 +220,8 @@ const agentsSchema = z.object({
   }),
   auto_escalate: z.boolean().default(false),
   max_delegation_depth: z.number().min(1).max(10).default(3),
+  /** Maximum tool-loop iterations before the agent stops. */
+  max_iterations: z.number().min(1).max(50).default(10),
 }).default({});
 
 const embeddingProviderSchema = z.enum(['openai', 'gemini', 'ollama', 'llamacpp', 'voyage']);