fix: set legacy skill field from skills list in normalize_job

deploy-crons normalize_job() was normalizing model/provider fields
but ignoring skill/skills. Jobs with a `skills` list but no `skill`
field would be stored without the legacy field set, bypassing the
normalization that _apply_skill_fields() in cron/jobs.py provides.

Now normalize_job() deduplicates and sets both `skills` (list) and
`skill` (first element) using the same logic as _apply_skill_fields().

Fixes #579

deploy-crons.py

@@ -18,9 +18,9 @@ from typing import Any, Dict, Optional
 
 def normalize_job(job: Dict[str, Any]) -> Dict[str, Any]:
     """
-    Normalize a job dict to ensure consistent model field types.
+    Normalize a job dict to ensure consistent model field types and aligned skill fields.
 
-    Before normalization:
+    Model normalization:
     - If model AND provider: model = raw string, provider = raw string (inconsistent)
     - If only model: model = raw string
     - If only provider: provider = raw string at top level
@@ -30,37 +30,61 @@ def normalize_job(job: Dict[str, Any]) -> Dict[str, Any]:
     - If provider exists: model = {"provider": "yyy"}
     - If both exist: model = {"model": "xxx", "provider": "yyy"}
     - If neither: model = None
+
+    Skill normalization:
+    - Aligns legacy `skill` (single string) with `skills` (list), setting skill = skills[0]
     """
     job = dict(job)  # Create a copy to avoid modifying the original
-    
+
+    # --- skill / skills normalization ---
+    raw_skill = job.get("skill")
+    raw_skills = job.get("skills")
+
+    if raw_skills is None:
+        skill_items = [raw_skill] if raw_skill else []
+    elif isinstance(raw_skills, str):
+        skill_items = [raw_skills]
+    else:
+        skill_items = list(raw_skills)
+
+    normalized_skills: list = []
+    for item in skill_items:
+        text = str(item or "").strip()
+        if text and text not in normalized_skills:
+            normalized_skills.append(text)
+
+    job["skills"] = normalized_skills
+    job["skill"] = normalized_skills[0] if normalized_skills else None
+
+    # --- model / provider normalization ---
     model = job.get("model")
     provider = job.get("provider")
-    
+
     # Skip if already normalized (model is a dict)
     if isinstance(model, dict):
         return job
-    
+
     # Build normalized model dict
     model_dict = {}
-    
+
     if model is not None and isinstance(model, str):
         model_dict["model"] = model.strip()
-    
+
     if provider is not None and isinstance(provider, str):
         model_dict["provider"] = provider.strip()
-    
+
     # Set model field
     if model_dict:
         job["model"] = model_dict
     else:
         job["model"] = None
-    
+
     # Remove top-level provider field if it was moved into model dict
     if provider is not None and "provider" in model_dict:
         # Keep provider field for backward compatibility but mark it as deprecated
         # This allows existing code that reads job["provider"] to continue working
         pass
-    
+
     return job
 
 
@@ -90,20 +114,26 @@ def normalize_jobs_file(jobs_file: Path, dry_run: bool = False) -> int:
     for i, job in enumerate(jobs):
         original_model = job.get("model")
         original_provider = job.get("provider")
-        
+        original_skill = job.get("skill")
+        original_skills = job.get("skills")
+
         normalized_job = normalize_job(job)
-        
+
         # Check if anything changed
         if (normalized_job.get("model") != original_model or
-            normalized_job.get("provider") != original_provider):
+                normalized_job.get("provider") != original_provider or
+                normalized_job.get("skill") != original_skill or
+                normalized_job.get("skills") != original_skills):
             jobs[i] = normalized_job
             modified_count += 1
-            
+
             job_id = job.get("id", "?")
             job_name = job.get("name", "(unnamed)")
             print(f"Normalized job {job_id} ({job_name}):")
             print(f"  model: {original_model!r} -> {normalized_job.get('model')!r}")
             print(f"  provider: {original_provider!r} -> {normalized_job.get('provider')!r}")
+            print(f"  skill: {original_skill!r} -> {normalized_job.get('skill')!r}")
+            print(f"  skills: {original_skills!r} -> {normalized_job.get('skills')!r}")
     
     if modified_count == 0:
         print("All jobs already have consistent model field types.")

tools/session_templates.py

@@ -1,271 +0,0 @@
-"""
-Session templates for code-first seeding.
-
-Research finding: Code-heavy sessions (execute_code dominant in first 30 turns)
-improve over time. File-heavy sessions degrade. Key is deterministic feedback loops.
-"""
-
-import json
-import logging
-import sqlite3
-import time
-from pathlib import Path
-from typing import Dict, List, Optional, Any
-from dataclasses import dataclass, asdict
-from enum import Enum
-
-logger = logging.getLogger(__name__)
-
-TEMPLATE_DIR = Path.home() / ".hermes" / "session-templates"
-
-
-class TaskType(Enum):
-    CODE = "code"
-    FILE = "file"
-    RESEARCH = "research"
-    MIXED = "mixed"
-
-
-@dataclass
-class ToolExample:
-    tool_name: str
-    arguments: Dict[str, Any]
-    result: str
-    success: bool
-    
-    def to_dict(self):
-        return asdict(self)
-    
-    @classmethod
-    def from_dict(cls, data):
-        return cls(**data)
-
-
-@dataclass
-class SessionTemplate:
-    name: str
-    task_type: TaskType
-    examples: List[ToolExample]
-    created_at: float = 0.0
-    usage_count: int = 0
-    
-    def __post_init__(self):
-        if self.created_at == 0.0:
-            self.created_at = time.time()
-    
-    def to_dict(self):
-        data = asdict(self)
-        data['task_type'] = self.task_type.value
-        return data
-    
-    @classmethod
-    def from_dict(cls, data):
-        data['task_type'] = TaskType(data['task_type'])
-        data['examples'] = [ToolExample.from_dict(e) for e in data.get('examples', [])]
-        return cls(**data)
-
-
-class SessionTemplates:
-    def __init__(self, template_dir=None):
-        self.template_dir = template_dir or TEMPLATE_DIR
-        self.template_dir.mkdir(parents=True, exist_ok=True)
-        self.templates = {}
-        self._load()
-    
-    def _load(self):
-        for f in self.template_dir.glob("*.json"):
-            try:
-                with open(f) as fh:
-                    data = json.load(fh)
-                    t = SessionTemplate.from_dict(data)
-                    self.templates[t.name] = t
-            except Exception as e:
-                logger.warning(f"Failed to load {f}: {e}")
-    
-    def _save(self, template):
-        path = self.template_dir / f"{template.name}.json"
-        with open(path, 'w') as f:
-            json.dump(template.to_dict(), f, indent=2)
-    
-    def classify(self, tool_calls):
-        if not tool_calls:
-            return TaskType.MIXED
-        
-        code = {'execute_code', 'code_execution'}
-        file_ops = {'read_file', 'write_file', 'patch', 'search_files'}
-        research = {'web_search', 'web_fetch', 'browser_navigate'}
-        
-        names = [tc.get('tool_name', '') for tc in tool_calls]
-        total = len(names)
-        
-        code_ratio = sum(1 for n in names if n in code) / total
-        file_ratio = sum(1 for n in names if n in file_ops) / total
-        research_ratio = sum(1 for n in names if n in research) / total
-        
-        if code_ratio > 0.6:
-            return TaskType.CODE
-        elif file_ratio > 0.6:
-            return TaskType.FILE
-        elif research_ratio > 0.6:
-            return TaskType.RESEARCH
-        return TaskType.MIXED
-    
-    def extract(self, session_id, max_examples=10):
-        db_path = Path.home() / ".hermes" / "state.db"
-        if not db_path.exists():
-            return []
-        
-        try:
-            conn = sqlite3.connect(str(db_path))
-            conn.row_factory = sqlite3.Row
-            
-            rows = conn.execute("""
-                SELECT role, content, tool_calls
-                FROM messages WHERE session_id = ?
-                ORDER BY timestamp LIMIT 100
-            """, (session_id,)).fetchall()
-            conn.close()
-            
-            examples = []
-            for row in rows:
-                if len(examples) >= max_examples:
-                    break
-                
-                if row['role'] == 'assistant' and row['tool_calls']:
-                    try:
-                        tcs = json.loads(row['tool_calls'])
-                        for tc in tcs:
-                            if len(examples) >= max_examples:
-                                break
-                            name = tc.get('function', {}).get('name')
-                            if not name:
-                                continue
-                            try:
-                                args = json.loads(tc.get('function', {}).get('arguments', '{}'))
-                            except:
-                                args = {}
-                            examples.append(ToolExample(name, args, "", True))
-                    except:
-                        continue
-                elif row['role'] == 'tool' and examples and examples[-1].result == "":
-                    examples[-1].result = row['content'] or ""
-            
-            return examples
-        except Exception as e:
-            logger.error(f"Extract failed: {e}")
-            return []
-    
-    def create(self, session_id, name=None, task_type=None, max_examples=10):
-        examples = self.extract(session_id, max_examples)
-        if not examples:
-            return None
-        
-        if task_type is None:
-            task_type = self.classify([{'tool_name': e.tool_name} for e in examples])
-        
-        if name is None:
-            name = f"{task_type.value}_{session_id[:8]}_{int(time.time())}"
-        
-        template = SessionTemplate(name, task_type, examples)
-        self.templates[name] = template
-        self._save(template)
-        logger.info(f"Created template {name} with {len(examples)} examples")
-        return template
-    
-    def get(self, task_type):
-        matching = [t for t in self.templates.values() if t.task_type == task_type]
-        if not matching:
-            return None
-        matching.sort(key=lambda t: t.usage_count)
-        return matching[0]
-    
-    def inject(self, template, messages):
-        if not template.examples:
-            return messages
-        
-        injection = [{
-            "role": "system",
-            "content": f"Template: {template.name} ({template.task_type.value})\nSuccessful tool call examples:"
-        }]
-        
-        for i, ex in enumerate(template.examples):
-            injection.append({
-                "role": "assistant",
-                "content": None,
-                "tool_calls": [{
-                    "id": f"tpl_{i}",
-                    "type": "function",
-                    "function": {"name": ex.tool_name, "arguments": json.dumps(ex.arguments)}
-                }]
-            })
-            injection.append({
-                "role": "tool",
-                "tool_call_id": f"tpl_{i}",
-                "content": ex.result
-            })
-        
-        idx = 0
-        for i, msg in enumerate(messages):
-            if msg.get("role") != "system":
-                break
-            idx = i + 1
-        
-        for i, msg in enumerate(injection):
-            messages.insert(idx + i, msg)
-        
-        template.usage_count += 1
-        self._save(template)
-        return messages
-    
-    def list(self, task_type=None):
-        templates = list(self.templates.values())
-        if task_type:
-            templates = [t for t in templates if t.task_type == task_type]
-        templates.sort(key=lambda t: t.created_at, reverse=True)
-        return templates
-    
-    def delete(self, name):
-        if name not in self.templates:
-            return False
-        del self.templates[name]
-        path = self.template_dir / f"{name}.json"
-        if path.exists():
-            path.unlink()
-        return True
-
-
-if __name__ == "__main__":
-    import argparse
-    parser = argparse.ArgumentParser()
-    sub = parser.add_subparsers(dest="cmd")
-    
-    list_p = sub.add_parser("list")
-    list_p.add_argument("--type", choices=["code", "file", "research", "mixed"])
-    
-    create_p = sub.add_parser("create")
-    create_p.add_argument("session_id")
-    create_p.add_argument("--name")
-    create_p.add_argument("--type", choices=["code", "file", "research", "mixed"])
-    create_p.add_argument("--max", type=int, default=10)
-    
-    delete_p = sub.add_parser("delete")
-    delete_p.add_argument("name")
-    
-    args = parser.parse_args()
-    ts = SessionTemplates()
-    
-    if args.cmd == "list":
-        tt = TaskType(args.type) if args.type else None
-        for t in ts.list(tt):
-            print(f"{t.name}: {t.task_type.value} ({len(t.examples)} examples, used {t.usage_count}x)")
-    elif args.cmd == "create":
-        tt = TaskType(args.type) if args.type else None
-        t = ts.create(args.session_id, args.name, tt, args.max)
-        if t:
-            print(f"Created: {t.name} ({len(t.examples)} examples)")
-        else:
-            print("Failed")
-    elif args.cmd == "delete":
-        print("Deleted" if ts.delete(args.name) else "Not found")
-    else:
-        parser.print_help()