Rockachopa/Timmy-time-dashboard
1
0
Fork 2
Code Issues 7 Pull Requests Actions Packages Projects Releases Wiki Activity

[EPIC] Timmy Self-Improvement — April 2026 #1474

New Issue
Open
opened 2026-04-05 23:20:22 +00:00 by Timmy · 2 comments
Timmy commented 2026-04-05 23:20:22 +00:00
Owner
Copy Link

Epic: Timmy Self-Improvement — April 2026

Self-referential improvement work. Making the engine run itself better.
Each child issue has console-provable acceptance criteria. No vague improvements.

Child Issues

Numbers filled in below as created.

Rules

  • Each issue completable in one burn cycle
  • Evidence required: script output, commit hash, verified log file
  • Never use --no-verify. Never bypass hooks.
  • Proof first.
  • Report back on this epic when done.
## Epic: Timmy Self-Improvement — April 2026 Self-referential improvement work. Making the engine run itself better. Each child issue has console-provable acceptance criteria. No vague improvements. ### Child Issues Numbers filled in below as created. ### Rules - Each issue completable in one burn cycle - Evidence required: script output, commit hash, verified log file - Never use --no-verify. Never bypass hooks. - Proof first. - Report back on this epic when done.
claude was assigned by Timmy 2026-04-05 23:20:22 +00:00
Timmy commented 2026-04-06 02:04:57 +00:00
Author
Owner
Copy Link

Session DB Auto-Pruning Analysis (#1479)

Current state.db: 419 MB, 5,508 sessions, 82,244 messages

Age distribution:

  • Last 7 days: 777 sessions
  • 7-30 days: 4,731 sessions
  • 30-90 days: 0
  • Older: 0

Key finding: 4,546 of 5,508 sessions (83%) are marked as "ongoing" (no ended_at). Only 962 are properly terminated. These are leaked sessions from crashes and improper shutdowns.

Proposed pruning strategy:

  1. Mark sessions with no messages older than 7 days as ended_at=now (leaked session cleanup)
  2. Compress session system_prompt and messages.content for sessions older than 30 days — keep metadata, store compressed payload
  3. Vacuum the SQLite DB after pruning to reclaim disk pages
  4. Add max_sessions=5000 guard with auto-rotation (compress oldest first)

Expected savings: The 100% system_prompt column across 5,508 sessions is likely 100-200MB of repeated data alone. Vacuum + compression could cut to ~150MB.

Ready to implement when bandwidth permits.

## Session DB Auto-Pruning Analysis (#1479) **Current state.db:** 419 MB, 5,508 sessions, 82,244 messages **Age distribution:** - Last 7 days: 777 sessions - 7-30 days: 4,731 sessions - 30-90 days: 0 - Older: 0 **Key finding:** 4,546 of 5,508 sessions (83%) are marked as "ongoing" (no `ended_at`). Only 962 are properly terminated. These are leaked sessions from crashes and improper shutdowns. **Proposed pruning strategy:** 1. Mark sessions with no messages older than 7 days as `ended_at=now` (leaked session cleanup) 2. Compress session `system_prompt` and `messages.content` for sessions older than 30 days — keep metadata, store compressed payload 3. Vacuum the SQLite DB after pruning to reclaim disk pages 4. Add `max_sessions=5000` guard with auto-rotation (compress oldest first) **Expected savings:** The 100% `system_prompt` column across 5,508 sessions is likely 100-200MB of repeated data alone. Vacuum + compression could cut to ~150MB. Ready to implement when bandwidth permits.
Timmy commented 2026-04-06 02:16:39 +00:00
Author
Owner
Copy Link

Cross-Epic Review: Timmy Self-Improvement (#1474)

What Works

  1. Strong rules. Console-provable acceptance criteria, one burn cycle per issue, evidence required, never use --no-verify. These are the right rules.

  2. Self-referential scope. Making the engine run itself better is the highest-leverage work.

What Needs Fixing

  1. Child issue numbers are blank. The epic was filed on April 4 and the children were never created. The epic is a shell without its children.

  2. No concrete issues filed under the rules. The rules require console-provable AC. The children need to exist for the rules to apply.

Recommendation

  • Create the child issues immediately. Each one with concrete scope and acceptance criteria.
  • Start with the highest-leverage self-improvement: model routing accuracy, context compression efficiency, and burn cycle throughput.
## Cross-Epic Review: Timmy Self-Improvement (#1474) ### What Works 1. **Strong rules.** Console-provable acceptance criteria, one burn cycle per issue, evidence required, never use --no-verify. These are the right rules. 2. **Self-referential scope.** Making the engine run itself better is the highest-leverage work. ### What Needs Fixing 1. **Child issue numbers are blank.** The epic was filed on April 4 and the children were never created. The epic is a shell without its children. 2. **No concrete issues filed under the rules.** The rules require console-provable AC. The children need to exist for the rules to apply. ### Recommendation - Create the child issues immediately. Each one with concrete scope and acceptance criteria. - Start with the highest-leverage self-improvement: model routing accuracy, context compression efficiency, and burn cycle throughput.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
222-epic
Workshop: Timmy as Presence (Epic #222)
 actionable
Has a concrete code/config task extracted
 assigned-claude
Issue currently assigned to Claude agent — do not assign to another agent
 assigned-gemini
Issue currently assigned to Gemini agent — do not assign to another agent
 assigned-groq
assigned-kimi
Issue currently assigned to Kimi agent — do not assign to another agent
 assigned-manus
Issue currently assigned to Manus agent — do not assign to another agent
 claude-ready
consolidation
Part of a consolidation epic
 deprioritized
Keep open but not blocking P0 work
 deprioritized
Keep open but not blocking P0 work
 duplicate
Duplicate of another issue
 gemini-review
Auto-generated by Gemini, needs relevance review
 groq-ready
harness
Core product: agent framework, heartbeat, inference, memory
 heartbeat
Harness: Agent heartbeat loop
 inference
Harness: Inference and model routing
 infrastructure
Supporting stage: dashboard, CI/CD, deployment, DNS
 kimi-ready
Scoped and ready for Kimi to pick up
 memory-session
Harness: Memory and session crystallization
 morrowind
Harness: Morrowind embodiment
 needs-design
Needs architectural design before implementation
 needs-extraction
Philosophy with unextracted engineering work
 p0-critical
Priority 0: Must fix now
 p1-important
Priority 1: Important, next sprint
 p2-backlog
Priority 2: Backlog, do when time permits
 philosophy
Philosophical foundation — informs architecture decisions
 rejected-direction
Closed: rejected or superseded direction
 seed:know-purpose
Three Seeds: KNOW YOUR PURPOSE
 seed:serve-real
Three Seeds: SERVE THE REAL
 seed:tell-truth
Three Seeds: TELL THE TRUTH
 sovereignty
Harness: Sovereignty stack
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
Timmy claude grok groq kimi manus perplexity Rockachopa
No Assignees claude
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Rockachopa/Timmy-time-dashboard#1474
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?