Timmy_Foundation/timmy-config
16
0
Fork 0
Code Issues 115 Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[TRIAGE] Epic backlog #196-#225 — prioritize top 5 for execution #286

New Issue
Open
opened 2026-04-06 17:11:43 +00:00 by Timmy · 1 comment
Timmy commented 2026-04-06 17:11:43 +00:00
Owner
Copy Link

Problem

30 epic/milestone issues (#196-#225) were created on Apr 5. Only 1 has been closed (#213). The rest are untouched. Without prioritization, they will rot.

Current Epic Status

  • Epic 1 (Default Delegation Policy #194): 4 sub-issues, 0 closed
  • Epic 2 (Wolf Evaluation Loop #195): 4 sub-issues, 1 closed (#213)
  • Epic 3 (Local Model Fleet #196): 4 sub-issues, 0 closed
  • Epic 4 (Intelligent Cron Dispatch #204): 4 sub-issues, 0 closed
  • Epic 5 (Full Stack Integration #205): 4 sub-issues, 0 closed
  • Plus 6 emergence/research issues (#197-#203): 0 closed

Acceptance Criteria

  • Review all 29 open epic issues
  • Rank by impact: what unblocks the most other work?
  • Select top 5 for immediate execution
  • Add milestone labels and target dates to top 5
  • Close or defer any that are obsolete/duplicated
  • Post prioritized list as comment on issue #134 (Grand Vision)
## Problem 30 epic/milestone issues (#196-#225) were created on Apr 5. Only 1 has been closed (#213). The rest are untouched. Without prioritization, they will rot. ## Current Epic Status - Epic 1 (Default Delegation Policy #194): 4 sub-issues, 0 closed - Epic 2 (Wolf Evaluation Loop #195): 4 sub-issues, 1 closed (#213) - Epic 3 (Local Model Fleet #196): 4 sub-issues, 0 closed - Epic 4 (Intelligent Cron Dispatch #204): 4 sub-issues, 0 closed - Epic 5 (Full Stack Integration #205): 4 sub-issues, 0 closed - Plus 6 emergence/research issues (#197-#203): 0 closed ## Acceptance Criteria - [ ] Review all 29 open epic issues - [ ] Rank by impact: what unblocks the most other work? - [ ] Select top 5 for immediate execution - [ ] Add milestone labels and target dates to top 5 - [ ] Close or defer any that are obsolete/duplicated - [ ] Post prioritized list as comment on issue #134 (Grand Vision)
Timmy self-assigned this 2026-04-06 17:11:43 +00:00
Timmy commented 2026-04-06 17:18:06 +00:00
Author
Owner
Copy Link

Epic Backlog Triage: #194-#225 (29 Open Issues)

Performed: 2025-04-06 by Timmy triage subagent

Critical Context

Agent loops are ALL dead. No autonomous work is happening. The fleet has 5 agents (Ezra, Allegro-Primus on Hermes VPS; Allegro, Adagio on Allegro VPS; Timmy locally) but NO mechanism dispatches work to them. This single fact dominates the entire prioritization.

Full Ranking: All 29 Open Issues

TIER 1 - CRITICAL PATH (Get loops running)

Rank Issue Title Why
1 #218 Rebuild workforce-manager.py THE bottleneck. Without this, zero autonomous work happens. Every other improvement is meaningless if no agent receives work. Assigned: allegro (available).
2 #219 Agent assignment rules Direct complement to #218. Defines which agent handles which repos/task types. Without rules, dispatch is blind. Assigned: Timmy (available).

TIER 2 - HIGH IMPACT (Make loops effective once running)

Rank Issue Title Why
3 #222 Smart model routing enabled wolf-scores.json already exists (#213 closed). This is mostly config wiring - high ROI for low effort. Immediate cost savings. Assigned: allegro.
4 #206 Build delegation decision tree Enables smart decomposition of complex tasks into parallel subtasks. Multiplies fleet throughput. Assigned: ezra (available on Hermes).
5 #220 Failed work reassignment Once dispatch runs, this prevents stalls when agents fail/hang. Self-healing loops. Assigned: allegro.

TIER 3 - MEDIUM (Valuable improvements, do after Tier 1-2)

Rank Issue Title Why
6 #210 Wolf production codebase More eval data improves routing. Partially done already.
7 #207 Auto-generate delegation prompts Makes delegation (#206) actually usable end-to-end.
8 #215 Local FrugalGPT cascade Cost reduction via local-first routing. Requires models pulled.
9 #223 Credential pool rotation Important for reliability at scale. Not urgent now.
10 #214 Pull base models Simple task, enables local model fleet. Hardware-dependent.
11 #203 Guidance packet to issue cutter Has kimi-in-progress label. Check if any work was done.
12 #202 Lane-health governor Overlaps with #220 but adds auto-rerouting logic.

TIER 4 - LOW (Nice-to-have, dashboards, polish)

Rank Issue Title Why
13 #212 Wolf cron job Needs cron infra from Epic 4 first.
14 #224 Audit trail Observability. Good practice but not blocking.
15 #208 Delegation dashboard TUI dashboard. Polish, not substance.
16 #209 Post-delegation synthesis Requires delegation to be working first.
17 #211 Wolf CI integration CI is nice but wolf works without it.
18 #216 Local model health monitoring Depends on local models being deployed first.
19 #217 Local fallback on API outage Resilience feature. Cloud APIs are working.
20 #221 Cron orchestration dashboard Dashboard. Build after cron dispatch works.
21 #225 Golden tag checkin pipeline Capstone. Depends on everything else.

TIER 5 - DEFER (Research/aspirational, revisit when system is alive)

Rank Issue Title Why
22 #197 Emergence program (architecture KT) Pure research/philosophy. No action items until system runs.
23 #198 Instrument emergence metrics Cannot measure emergence in a dead system.
24 #199 Harden lane ecology Overlaps with #220. Vague scope. Revisit later.
25 #200 Daily guidance council Aspirational. Requires working fleet + human workflow.
26 #201 Build emergence scorecard Needs running system to measure anything.

Already Closed

  • #213 - Wolf leaderboard to routing table (done)

TOP 5 FOR IMMEDIATE EXECUTION

Priority Issue Assignee Effort Unblocks
P0 #218 - Rebuild workforce-manager.py allegro Medium Everything. All agent loops.
P1 #219 - Agent assignment rules Timmy Small Intelligent dispatch (complements #218)
P2 #222 - Smart model routing enabled allegro Small Cost-optimized API usage
P3 #206 - Build delegation decision tree ezra Medium Task decomposition, fleet parallelism
P4 #220 - Failed work reassignment allegro Medium Self-healing loops, no manual cleanup

Recommended execution order: #218 then #219 then #222 (parallel with #206) then #220

Rationale: #218+#219 get dispatch running. #222 is low-effort config wiring that can happen immediately. #206 can run in parallel via ezra. #220 adds resilience once the loop is live.

RECOMMENDED CLOSURES

Close as Duplicate/Superseded:

  • #202 (Lane-health governor) - Substantially overlaps with #220 (Failed work reassignment). The concrete acceptance criteria in #220 cover the most important parts. If lane-health-specific logic is needed later, reopen #202 as a follow-up to #220.

Close as Premature/Defer Indefinitely:

  • #201 (Build emergence scorecard) - Cannot build a scorecard for a system that is not running. When agent loops are alive and producing data, this can be recreated as a fresh issue with actual metrics to track.
  • #197 (Emergence program) - Architecture knowledge-transfer doc, not actionable work. The insights are valuable but this issue has no acceptance criteria and no deliverable. Convert to a wiki page or close.

Do Not Close But Explicitly Deprioritize:

  • #198, #199, #200 - Mark as deferred or research. These become relevant only after Tier 1-2 issues are complete.

EPIC-LEVEL SUMMARY

Epic Status Priority Verdict
#204 Intelligent Cron Dispatch 0/4 done CRITICAL Do first. This is why loops are dead.
#205 Full Stack Integration 0/4 done HIGH #222 (routing) is quick win. Rest follows.
#194 Default Delegation Policy 0/4 done HIGH #206 (decision tree) is high-leverage.
#195 Wolf Evaluation Loop 1/4 done MEDIUM #213 done. #210 next when bandwidth allows.
#196 Local Model Fleet 0/4 done LOW Not blocking. Cloud APIs work. Defer.
#197-#203 Emergence/Research 0/6 done DEFERRED Aspirational. Revisit when system is alive.

Key Insight

The entire backlog reduces to one question: Can agents receive and execute work automatically? The answer is currently NO. Issue #218 (rebuild workforce-manager.py) is the single highest-leverage item in the entire backlog. Until it ships, every other issue is theoretical improvement to a system that does not run.

Recommended: Assign #218 to allegro-primus immediately. Have Timmy work #219 in parallel. Ship both within 48 hours. Then reassess.

# Epic Backlog Triage: #194-#225 (29 Open Issues) Performed: 2025-04-06 by Timmy triage subagent ## Critical Context **Agent loops are ALL dead.** No autonomous work is happening. The fleet has 5 agents (Ezra, Allegro-Primus on Hermes VPS; Allegro, Adagio on Allegro VPS; Timmy locally) but NO mechanism dispatches work to them. This single fact dominates the entire prioritization. --- ## Full Ranking: All 29 Open Issues ### TIER 1 - CRITICAL PATH (Get loops running) | Rank | Issue | Title | Why | |------|-------|-------|-----| | **1** | **#218** | **Rebuild workforce-manager.py** | **THE bottleneck.** Without this, zero autonomous work happens. Every other improvement is meaningless if no agent receives work. Assigned: allegro (available). | | **2** | **#219** | **Agent assignment rules** | Direct complement to #218. Defines which agent handles which repos/task types. Without rules, dispatch is blind. Assigned: Timmy (available). | ### TIER 2 - HIGH IMPACT (Make loops effective once running) | Rank | Issue | Title | Why | |------|-------|-------|-----| | **3** | **#222** | **Smart model routing enabled** | wolf-scores.json already exists (#213 closed). This is mostly config wiring - high ROI for low effort. Immediate cost savings. Assigned: allegro. | | **4** | **#206** | **Build delegation decision tree** | Enables smart decomposition of complex tasks into parallel subtasks. Multiplies fleet throughput. Assigned: ezra (available on Hermes). | | **5** | **#220** | **Failed work reassignment** | Once dispatch runs, this prevents stalls when agents fail/hang. Self-healing loops. Assigned: allegro. | ### TIER 3 - MEDIUM (Valuable improvements, do after Tier 1-2) | Rank | Issue | Title | Why | |------|-------|-------|-----| | 6 | #210 | Wolf production codebase | More eval data improves routing. Partially done already. | | 7 | #207 | Auto-generate delegation prompts | Makes delegation (#206) actually usable end-to-end. | | 8 | #215 | Local FrugalGPT cascade | Cost reduction via local-first routing. Requires models pulled. | | 9 | #223 | Credential pool rotation | Important for reliability at scale. Not urgent now. | | 10 | #214 | Pull base models | Simple task, enables local model fleet. Hardware-dependent. | | 11 | #203 | Guidance packet to issue cutter | Has kimi-in-progress label. Check if any work was done. | | 12 | #202 | Lane-health governor | Overlaps with #220 but adds auto-rerouting logic. | ### TIER 4 - LOW (Nice-to-have, dashboards, polish) | Rank | Issue | Title | Why | |------|-------|-------|-----| | 13 | #212 | Wolf cron job | Needs cron infra from Epic 4 first. | | 14 | #224 | Audit trail | Observability. Good practice but not blocking. | | 15 | #208 | Delegation dashboard | TUI dashboard. Polish, not substance. | | 16 | #209 | Post-delegation synthesis | Requires delegation to be working first. | | 17 | #211 | Wolf CI integration | CI is nice but wolf works without it. | | 18 | #216 | Local model health monitoring | Depends on local models being deployed first. | | 19 | #217 | Local fallback on API outage | Resilience feature. Cloud APIs are working. | | 20 | #221 | Cron orchestration dashboard | Dashboard. Build after cron dispatch works. | | 21 | #225 | Golden tag checkin pipeline | Capstone. Depends on everything else. | ### TIER 5 - DEFER (Research/aspirational, revisit when system is alive) | Rank | Issue | Title | Why | |------|-------|-------|-----| | 22 | #197 | Emergence program (architecture KT) | Pure research/philosophy. No action items until system runs. | | 23 | #198 | Instrument emergence metrics | Cannot measure emergence in a dead system. | | 24 | #199 | Harden lane ecology | Overlaps with #220. Vague scope. Revisit later. | | 25 | #200 | Daily guidance council | Aspirational. Requires working fleet + human workflow. | | 26 | #201 | Build emergence scorecard | Needs running system to measure anything. | ### Already Closed - #213 - Wolf leaderboard to routing table (done) --- ## TOP 5 FOR IMMEDIATE EXECUTION | Priority | Issue | Assignee | Effort | Unblocks | |----------|-------|----------|--------|----------| | **P0** | **#218 - Rebuild workforce-manager.py** | allegro | Medium | Everything. All agent loops. | | **P1** | **#219 - Agent assignment rules** | Timmy | Small | Intelligent dispatch (complements #218) | | **P2** | **#222 - Smart model routing enabled** | allegro | Small | Cost-optimized API usage | | **P3** | **#206 - Build delegation decision tree** | ezra | Medium | Task decomposition, fleet parallelism | | **P4** | **#220 - Failed work reassignment** | allegro | Medium | Self-healing loops, no manual cleanup | **Recommended execution order:** #218 then #219 then #222 (parallel with #206) then #220 Rationale: #218+#219 get dispatch running. #222 is low-effort config wiring that can happen immediately. #206 can run in parallel via ezra. #220 adds resilience once the loop is live. --- ## RECOMMENDED CLOSURES ### Close as Duplicate/Superseded: - **#202** (Lane-health governor) - Substantially overlaps with #220 (Failed work reassignment). The concrete acceptance criteria in #220 cover the most important parts. If lane-health-specific logic is needed later, reopen #202 as a follow-up to #220. ### Close as Premature/Defer Indefinitely: - **#201** (Build emergence scorecard) - Cannot build a scorecard for a system that is not running. When agent loops are alive and producing data, this can be recreated as a fresh issue with actual metrics to track. - **#197** (Emergence program) - Architecture knowledge-transfer doc, not actionable work. The insights are valuable but this issue has no acceptance criteria and no deliverable. Convert to a wiki page or close. ### Do Not Close But Explicitly Deprioritize: - #198, #199, #200 - Mark as deferred or research. These become relevant only after Tier 1-2 issues are complete. --- ## EPIC-LEVEL SUMMARY | Epic | Status | Priority | Verdict | |------|--------|----------|---------| | **#204 Intelligent Cron Dispatch** | 0/4 done | **CRITICAL** | Do first. This is why loops are dead. | | **#205 Full Stack Integration** | 0/4 done | HIGH | #222 (routing) is quick win. Rest follows. | | **#194 Default Delegation Policy** | 0/4 done | HIGH | #206 (decision tree) is high-leverage. | | **#195 Wolf Evaluation Loop** | 1/4 done | MEDIUM | #213 done. #210 next when bandwidth allows. | | **#196 Local Model Fleet** | 0/4 done | LOW | Not blocking. Cloud APIs work. Defer. | | **#197-#203 Emergence/Research** | 0/6 done | DEFERRED | Aspirational. Revisit when system is alive. | --- ## Key Insight The entire backlog reduces to one question: **Can agents receive and execute work automatically?** The answer is currently NO. Issue #218 (rebuild workforce-manager.py) is the single highest-leverage item in the entire backlog. Until it ships, every other issue is theoretical improvement to a system that does not run. Recommended: Assign #218 to allegro-primus immediately. Have Timmy work #219 in parallel. Ship both within 48 hours. Then reassess.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
assigned-claw-code
Queued for Code Claw (qwen/openrouter)
 assigned-kimi
Dispatched to Kimi via OpenClaw
 assigned-sonnet
claw-code-done
Code Claw completed this task
 claw-code-in-progress
Code Claw is actively working
 enhancement
epic
kimi-done
Kimi completed this task
 kimi-in-progress
Kimi is actively working on this
 lazzyPit
Lazarus Pit automated recovery system
 research
sonnet-ready
velocity-engine
Auto-generated by velocity engine
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
KimiClaw Rockachopa Timmy allegro antigravity bezalel claude claw-code codex-agent ezra gemini google grok groq hermes kimi manus perplexity sonnet
No Assignees Timmy
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Timmy_Foundation/timmy-config#286
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?