Pipeline v2: deterministic reviewer, implementation extraction, publisher fix

- Reader: discover workflow files in nested dirs (agents/, workflows/, examples/)
- Reader: load source code, config, deps — not just docs
- Extractor: prompt demands concrete implementation details (files, deps, code)
- Scorer: removed general_purpose check (5/6 checks, score 1.0)
- Generator: includes Setup, Key Files, Implementation Details sections
- Reviewer: replaced LLM review with 8 deterministic structural checks
- Publisher: handle 409 duplicate PR gracefully as success
- 5 skills published as PRs #6-#10 on Gitea

config/settings.yaml

@@ -16,18 +16,31 @@ llm:
   api_key: ""
   max_tokens: 8000
 
+# Secondary LLM for pipeline tasks — uses Ollama on 3060 (non-reasoning model)
+llm_pipeline:
+  base_url: http://100.64.0.4:11434
+  model: qwen2.5:7b
+  api_key: ""
+  max_tokens: 6000
+
 scout:
   queries:
-    - 'agent framework langgraph mcp multi-agent'
-    - 'ai workflow agent pipeline rag pipeline'
-    - 'llm orchestration tool-use tool calling'
+    - 'langchain workflow example'
+    - 'langgraph agent workflow'
+    - 'autogen multi-agent example'
+    - 'crewai task workflow'
+    - 'llamaindex pipeline example'
+    - 'mcp server implementation'
+    - 'rag agent workflow'
+    - 'tool calling workflow'
   filters:
-    stars_min: 10
-    pushed_after: 2026-06-01
+    stars_min: 15
+    pushed_after: 2026-05-01
     language: Python
     archived: false
-  max_results: 30
-  cooldown_hours: 24
+    size_max_kb: 10000
+  max_results: 15
+  cooldown_hours: 6
 
 filter:
   categories: