refactor: replace orchestrator/verifier chain with direct LiteLLM calls

Drop the three-layer Claude subprocess orchestration (local model →
Claude verifier → cloud escalation). Skills now call LiteLLM directly
and return plain text to Claude Code, which decides what to do with it.

- Delete executor, orchestrator, verifier, result, attempts packages
- Simplify LiteLLMExecutor: Run(Request)→Result becomes Complete(model,sys,user)→(string,int64,error)
- Replace ExecutorFn with CompleteFunc in all 6 skill configs
- Rewrite all skill handlers to call Complete and return {"text","model","duration_ms"}
- Simplify config/models: remove Verifier/LlamaSwapURL, add ModelFor
- Bump version to v0.5.0

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

internal/skills/retrospective/skill.go

@@ -5,19 +5,18 @@ import (
 	"context"
 	"encoding/json"
 
-	iexec "github.com/mathiasbq/supervisor/internal/exec"
 	"github.com/mathiasbq/supervisor/internal/registry"
 )
 
-// ExecutorFn allows injecting a test double for the subprocess executor.
-type ExecutorFn func(ctx context.Context, req iexec.Request) (iexec.Result, error)
+// CompleteFunc is the function used to call a local model.
+type CompleteFunc func(ctx context.Context, model, system, user string) (string, int64, error)
 
 // Config holds retrospective skill configuration.
 type Config struct {
-	SkillPrompt  string     // content of retrospective.md
-	DefaultModel string     // model to use when not specified in args
-	SessionsDir  string     // path to brain/sessions/
-	ExecutorFn   ExecutorFn // injected executor
+	SkillPrompt  string
+	DefaultModel string
+	SessionsDir  string
+	CompleteFunc CompleteFunc
 }
 
 // Skill implements registry.Skill for the retrospective tool.
@@ -36,7 +35,7 @@ func (s *Skill) Tools() []registry.ToolDef {
 	return []registry.ToolDef{
 		{
 			Name:        "retrospective",
-			Description: "Run a retrospective on a completed session. Reads the session log, identifies novel learnings, and writes structured entries to the brain for ingestion. Call at the end of each coding session.",
+			Description: "Consult a local model to analyse a completed session and identify what is novel or worth preserving as organizational knowledge.",
 			InputSchema: json.RawMessage(`{
 				"type": "object",
 				"required": ["session_id"],