Skip to content

seed(elixir-oban-worker): full 12-dimension run with real competition#31

Merged
ty13r merged 1 commit intomainfrom
feat/seed-run-oban-worker
Apr 12, 2026
Merged

seed(elixir-oban-worker): full 12-dimension run with real competition#31
ty13r merged 1 commit intomainfrom
feat/seed-run-oban-worker

Conversation

@ty13r
Copy link
Copy Markdown
Owner

@ty13r ty13r commented Apr 12, 2026

Summary

  • Seed run v2.0 Phase 3: Variant evolution orchestrator + Scientist + Spawner #4: elixir-oban-worker with 12 dimensions (1 foundation + 11 capabilities)
  • Full pipeline, no shortcuts: 12 Spawner + 48 Competitor + 1 Engineer = 61 Opus dispatches
  • Both variants competed against 2 challenges per dimension (24 challenges total)
  • Mean winning fitness: 0.487

Competition results (real head-to-head)

Dimension V1 (seed) V2 (spawn) Winner Delta
worker-philosophy (F) 0.245 0.265 v2 +0.02
perform-callback 0.140 0.140 tie 0.00
args-serialization 0.393 0.393 tie 0.00
retry-strategy 0.379 0.476 v2 +0.10
unique-constraints 0.668 0.668 tie 0.00
cron-scheduling 0.388 0.671 v2 +0.28
return-values 0.264 0.429 v2 +0.17
queues-and-priority 0.547 0.547 tie 0.00
testing-workers 0.398 0.234 V1 -0.16
telemetry 0.834 0.834 tie 0.00
transactional-jobs 0.637 0.261 V1 -0.38
recurring-jobs 0.301 0.387 v2 +0.09

v1 wins: 2 | v2 wins: 5 | ties: 5

Test plan

  • Both variants scored on every challenge (no synthetic data)
  • competition_scores matches CompetitionScoresPayload shape (12 matches, baseline_ran=true)
  • DB nuke + reboot loads correctly (25 genomes, 12 variants, 12 vevos, 24 challenges)
  • learning_log has 4 entries (created, finalized, integration_report, competition_scores)

🤖 Generated with Claude Code

Full pipeline seed run #4 — no shortcuts. 12 Spawner + 48 Competitor
dispatches + 1 Engineer. Both v1 (seed) and v2 (spawn) competed
against 2 challenges per dimension (24 challenges total, balanced
medium/hard).

Real competition results:
- v1 wins 2 dims (testing-workers 0.40>0.23, transactional-jobs 0.64>0.26)
- v2 wins 5 dims (retry-strategy, cron-scheduling, return-values,
  recurring-jobs, worker-philosophy)
- 5 ties (perform, args, unique, queues, telemetry)
- Mean winning fitness: 0.487

Composite: 619 lines, 12 capability sections, 3 cross-cutting examples,
10 common mistakes. Foundation: transactional-saga philosophy.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
@ty13r ty13r merged commit 6ff6c86 into main Apr 12, 2026
@ty13r ty13r deleted the feat/seed-run-oban-worker branch April 19, 2026 20:21
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant