Files
Chronicler e9c6034349 docs(codex): Add Phase 2 overview with 6-workspace plan
- 6 workspaces planned (Operations, Public KB, Subscriber KB, Brainstorming, Relationship, Pokerole Project)
- 3 accounts: mkrause612 (exists), gingerfury (Meg/Admin), Unicorn20089 (Holly/Staff)
- Holly gets staff access (Pokerole, Operations, Public KB, Subscriber KB)
- Ready to execute when Michael returns from rest
- Estimated: 35 minutes for workspace setup + account creation
2026-02-21 02:04:40 +00:00

8.9 KiB

🚀 Firefrost Codex — Phase 2 Overview

Date: February 20, 2026
Status: Phase 1 COMPLETE | Phase 2 READY TO EXECUTE
Current State: Operational AI assistant at $0/month with 5-10 second responses


WHAT WE'VE ACCOMPLISHED (Phase 1)

Infrastructure Deployed:

  • AnythingLLM + Ollama running on TX1 (Docker containers)
  • 5 models downloaded (73.5 GB total)
  • Primary model selected: qwen2.5-coder:7b (fast, quality responses)
  • Multi-user mode enabled
  • Admin account created (mkrause612)
  • Web interface accessible: http://38.68.14.26:3001
  • $0/month cost validated
  • ~10,100 lines of documentation created

Performance Validated:

  • Response time: 5-10 seconds
  • Quality: Good enough for community support
  • Resource usage: Minimal impact on game servers
  • Sustainability: No recurring costs

🎯 WHAT WE STILL NEED TO DO (Phase 2)

Priority 1: Workspace Setup (30 minutes, LOW RISK)

Current State: Only "default" workspace exists

What We Need:

  • Rename "default" → "Operations"
  • Create "Public KB" workspace (for marketing widget)
  • Create "Subscriber KB" workspace (for $1+ tier)
  • Create "Brainstorming" workspace (Michael/Meg ideation)
  • Create "Relationship" workspace (Chronicler continuity docs)

Assign Models:

  • Operations, Public KB, Subscriber KB → qwen2.5-coder:7b (fast)
  • Brainstorming → llama3.3:70b (deep reasoning)
  • Relationship → qwen2.5-coder:7b (fast)

Why It Matters:

  • Role-based access control (public vs subscriber vs staff)
  • Different knowledge bases for different audiences
  • Separates concerns (support vs ideation vs documentation)

Estimated Time: 30 minutes
Risk Level: Low (just UI configuration)


Priority 2: Test Document Upload (30 minutes, VALIDATION)

Current State: No documents uploaded yet

What We Need:

  • Test upload 3-5 documents from operations manual
  • Verify search works
  • Verify retrieval works
  • Test vector embeddings functionality
  • Identify any issues before bulk upload

Test Documents (suggested):

  • docs/core/infrastructure-manifest.md
  • docs/core/tasks.md
  • docs/standards/FFG-STD-001-revision-control.md
  • SESSION-HANDOFF-PROTOCOL.md
  • README.md

Why It Matters:

  • Validates document processing works
  • Catches issues early (before uploading hundreds of files)
  • Tests search quality with real operations docs
  • Ensures vector embeddings are working

Estimated Time: 30 minutes
Risk Level: Low (testing only, no production impact)


Priority 3: Create Meg's Account (15 minutes, IMPORTANT)

Current State: Only Michael (mkrause612) can log in

What We Need:

  • Create account: gingerfury
  • Set role: Admin
  • Grant access to all workspaces
  • Test login functionality
  • Verify she can use Codex

Why It Matters:

  • Meg is co-owner, needs full access
  • Partnership equality (both partners have admin)
  • Community management use case (Meg can use for support)

Estimated Time: 15 minutes
Risk Level: Very Low (just user creation)


Priority 4: Git Sync Process (1-2 hours, AUTOMATION)

Current State: Manual document upload only

What We Need:

  • Build script to sync Git repos → Codex workspaces
  • Map documents to correct workspaces:
    • Operations manual → Operations workspace
    • Public guides → Public KB workspace
    • Standards/SOPs → Subscriber KB workspace
    • Relationship docs → Relationship workspace
  • Test sync functionality
  • Document manual process (fallback)
  • Schedule automation (optional - can stay manual)

Two Approaches:

  1. Automated Script: Cron job syncs Git → Codex automatically
  2. Manual Process: Simple command to sync on-demand

Why It Matters:

  • Keeps Codex knowledge current with Git
  • Prevents documentation drift
  • Reduces manual maintenance
  • Ensures AI has latest ops manual content

Estimated Time: 1-2 hours (script) OR 30 min (manual process)
Risk Level: Medium (needs testing, but not critical path)


Priority 5: Security Hardening (2-3 hours, CRITICAL)

Current State: HTTP only, port 3001 exposed, no firewall rules

What We Need:

A. SSL/TLS Certificate (1 hour)

  • Install Nginx reverse proxy on TX1
  • Configure domain: codex.firefrostgaming.com
  • Get Let's Encrypt SSL certificate
  • Configure HTTPS redirect (HTTP → HTTPS)
  • Test certificate renewal

B. Firewall Configuration (30 minutes)

  • Install/configure UFW on TX1
  • Restrict port 3001 to localhost only
  • Allow port 80/443 (Nginx)
  • Allow game server ports
  • Test connections

C. Backup Automation (1 hour)

  • Test backup script (already exists)
  • Schedule daily backups (cron)
  • Configure retention policy (7 days local, 30 days remote?)
  • Test restoration process
  • Document backup/restore procedures

Why It Matters:

  • HTTPS prevents credential interception
  • Firewall limits attack surface
  • Backups prevent data loss
  • Professional security posture

Estimated Time: 2-3 hours
Risk Level: Medium-High (production service, needs careful testing)


Priority 6: Documentation Updates (30 minutes, HOUSEKEEPING)

What We Need:

  • Update tasks.md (mark Codex as Phase 2 in progress)
  • Update infrastructure-manifest.md (add Codex services)
  • Create user guide for Michael/Meg
  • Document workspace purposes
  • Update session handoff with Phase 2 status

Why It Matters:

  • Future Chroniclers need current state
  • Operations manual stays accurate
  • Prevents documentation drift

Estimated Time: 30 minutes
Risk Level: Very Low (just documentation)


📊 PHASE 2 TIMELINE ESTIMATES

Minimum Viable (Core Functionality):

  • Workspace setup: 30 min
  • Test documents: 30 min
  • Meg's account: 15 min
  • Total: ~1.5 hours

Recommended (Production Ready):

  • Above + Git sync (manual): 30 min
  • Above + Security (SSL only): 1 hour
  • Total: ~3 hours

Complete (Fully Hardened):

  • All items above
  • Full automation (Git sync script)
  • Complete security hardening (SSL + firewall + backups)
  • Total: ~6-7 hours

🎯 SUGGESTED EXECUTION ORDER

Session 1 (Today? 1.5-3 hours)

  1. Workspace Creation (30 min) - Low risk, high value
  2. Test Document Upload (30 min) - Validate before bulk work
  3. Meg's Account (15 min) - Quick win
  4. SSL/TLS Setup (1 hour) - Security baseline
  5. Commit & document progress

Session 2 (Future, 2-3 hours)

  1. Git Sync Process (1-2 hours) - Automation or manual
  2. Firewall Hardening (30 min) - Lock down access
  3. Backup Testing (1 hour) - Validate disaster recovery
  4. Final documentation (30 min)

Session 3 (Future, optional polish)

  1. Performance tuning
  2. Advanced features
  3. Marketing preparation
  4. Community beta testing

SUCCESS CRITERIA

Phase 2 is COMPLETE when:

5 workspaces exist with proper names and model assignments
Operations manual docs are uploaded and searchable
Meg can log in as gingerfury with admin access
Git sync process exists (automated OR manual, both acceptable)
HTTPS is working via codex.firefrostgaming.com
Firewall restricts access appropriately
Backups are scheduled and tested
All documentation is updated
Michael and Meg can both use Codex for real work


💰 COST IMPACT

Phase 2 Additional Costs: $0/month

Everything is self-hosted infrastructure we already own:

  • SSL certificate: Free (Let's Encrypt)
  • Nginx: Free (open source)
  • Backups: Free (to Command Center or existing storage)
  • Git sync: Free (bash script)

No new recurring costs. Budget constraint respected.


⚠️ RISKS & MITIGATION

Risk 1: SSL Configuration Breaks Web Access

Mitigation: Keep HTTP working during SSL setup, only redirect after testing
Rollback: Disable Nginx, access via :3001 directly

Risk 2: Document Upload Fails or Slow

Mitigation: Start with small test batch, scale gradually
Rollback: Remove bad documents, try different formats

Risk 3: Firewall Locks Us Out

Mitigation: Test each rule before applying, keep SSH access open
Rollback: UFW disable, revert rules

Risk 4: Backup Automation Fails

Mitigation: Test restore before trusting backups
Fallback: Manual backups, Docker volume preservation

None of these risks are catastrophic. Phase 1 works. Phase 2 is enhancement.


🚀 READY TO PROCEED?

Current Status: Firefrost Codex is operational and usable RIGHT NOW.

Phase 2 makes it:

  • More organized (workspaces)
  • More accessible (Meg's account)
  • More sustainable (Git sync)
  • More secure (HTTPS + firewall)
  • More resilient (backups)

But the core value is already delivered: $0/month AI assistant with 5-10 second responses.

What would you like to tackle first?


💙🔥❄️🚀

Fire + Frost + Foundation + Codex = Where Love Builds Legacy