chore: update claude-api skill [auto-sync] (#730)

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

skills/claude-api/python/agent-sdk/README.md

@@ -215,6 +215,16 @@ async for message in query(
         session_id = message.data.get("session_id")  # Capture for resuming later
 ```
 
+`AssistantMessage` includes per-turn `usage` data (a dict matching the Anthropic API usage shape) for tracking costs:
+
+```python
+from claude_agent_sdk import query, ClaudeAgentOptions, AssistantMessage
+
+async for message in query(prompt="...", options=ClaudeAgentOptions()):
+    if isinstance(message, AssistantMessage) and message.usage:
+        print(f"Input: {message.usage['input_tokens']}, Output: {message.usage['output_tokens']}")
+```
+
 Typed task message subclasses are available for better type safety when handling subagent task events:
 - `TaskStartedMessage` — emitted when a subagent task is registered
 - `TaskProgressMessage` — real-time progress updates with cumulative usage metrics

skills/claude-api/python/claude-api/README.md

@@ -111,7 +111,7 @@ response = client.messages.create(
 
 ## Prompt Caching
 
-Cache large context to reduce costs (up to 90% savings).
+Cache large context to reduce costs (up to 90% savings). **Caching is a prefix match** — any byte change anywhere in the prefix invalidates everything after it. For placement patterns, architectural guidance (frozen system prompt, deterministic tool order, where to put volatile content), and the silent-invalidator audit checklist, read `shared/prompt-caching.md`.
 
 ### Automatic Caching (Recommended)
 
@@ -156,6 +156,16 @@ response = client.messages.create(
 )
 ```
 
+### Verifying Cache Hits
+
+```python
+print(response.usage.cache_creation_input_tokens)  # tokens written to cache (~1.25x cost)
+print(response.usage.cache_read_input_tokens)      # tokens served from cache (~0.1x cost)
+print(response.usage.input_tokens)                 # uncached tokens (full cost)
+```
+
+If `cache_read_input_tokens` is zero across repeated identical-prefix requests, a silent invalidator is at work — `datetime.now()` or a UUID in the system prompt, unsorted `json.dumps()`, or a varying tool set. See `shared/prompt-caching.md` for the full audit table.
+
 ---
 
 ## Extended Thinking