Bootstrap torch environment and implement scheduler

- Created `torch/` directory structure.
- Created `torch/package.json` with locking and linting scripts.
- Created `torch/torch-config.json` with scheduler configuration.
- Created `torch/src/prompts/roster.json` and dummy prompts.
- Implemented `torch/scheduler.py` to handle daily agent execution flow:
    - Exclusion parsing
    - Round-robin selection
    - Locking with retry
    - Memory workflow (retrieve/store)
    - Validation (flake8)
    - Completion logging

Co-authored-by: PR0M3TH3AN <[email protected]>

torch/src/prompts/daily/ci-health-agent.md

@@ -0,0 +1,3 @@
+# CI Health Agent
+
+This agent runs CI checks.

torch/src/prompts/daily/content-audit-agent.md

@@ -0,0 +1,3 @@
+# Content Audit Agent
+
+This agent audits content.

torch/src/prompts/roster.json

@@ -0,0 +1,6 @@
+{
+  "daily": [
+    "ci-health-agent",
+    "content-audit-agent"
+  ]
+}