What does this PR do?

Adds a validation framework to altimate-code that evaluates AI agent traces against quality criteria (Groundedness, Validity, Coherence, Utility, Tool Validation).

• Adds /validate skill with SKILL.md and batch_validate.py that calls the Altimate backend API directly via SSE (no Langfuse dependency) for single trace, date range, and session modes
• Adds validate CLI command (altimate-code validate install/status) that installs the skill to ~/.altimate-code/skills/validate/
• Adds conversation-logger that auto-logs user/assistant turns to the backend on session idle
• Wires initConversationLogger() into InstanceBootstrap
• Auto-installs validate skill after bun run build
• Fixes build.ts to embed skill assets from the correct path
• Skill output is written directly to markdown files instead of terminal; user is notified with the file path after each write

Type of change

• New feature (non-breaking change which adds functionality)

How did you verify your code works?

• Ran /validate <trace_id> locally and confirmed per-trace markdown files are written to the report directory
• Verified altimate-code validate install correctly installs the skill
• Confirmed conversation logger attaches and logs turns on session idle
• Typecheck passes across all packages

Issue for this PR

N/A — internal feature addition

Checklist

• Tested locally and confirmed the feature works end-to-end
• No unrelated changes included in this PR