feat: add /add-ollama skill for local model inference (#712)

* feat: add /add-ollama skill for local model inference

Adds a skill that integrates Ollama as an MCP server, allowing the
container agent to offload tasks to local models (summarization,
translation, general queries) while keeping Claude as orchestrator.

Skill contents:
- ollama-mcp-stdio.ts: stdio MCP server with ollama_list_models and
  ollama_generate tools
- ollama-watch.sh: macOS notification watcher for Ollama activity
- Modifications to index.ts (MCP config) and container-runner.ts
  (log surfacing)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* chore: rename skill from /add-ollama to /add-ollama-tool

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: gavrielc <gabicohen22@yahoo.com>

.claude/skills/add-ollama-tool/modify/src/container-runner.ts.intent.md

@@ -0,0 +1,18 @@
+# Intent: src/container-runner.ts modifications
+
+## What changed
+Surface Ollama MCP server log lines at info level so they appear in `nanoclaw.log` for the monitoring watcher script.
+
+## Key sections
+
+### container.stderr handler (inside runContainerAgent)
+- Changed: empty line check from `if (line)` to `if (!line) continue;`
+- Added: `[OLLAMA]` tag detection — lines containing `[OLLAMA]` are logged at `logger.info` instead of `logger.debug`
+- All other stderr lines remain at `logger.debug` level
+
+## Invariants (must-keep)
+- Stderr truncation logic unchanged
+- Timeout reset logic unchanged (stderr doesn't reset timeout)
+- Stdout parsing logic unchanged
+- Volume mount logic unchanged
+- All other container lifecycle unchanged