Karpathy's auto-research is now a native Claude Code feature — /goal and /loop. The real unlock: evaluator cartridges — custom scoring algorithms that grade AI output autonomously so the loop knows when to stop. No human in the loop. We have zero of this.
A cartridge is a self-contained scoring algorithm: text in → score out (1–10). Composite across multiple cartridges for holistic grading. The newsletter loop ran 5 attempts autonomously — 7.1 → 26.8 over 40 minutes. Zero keystrokes from the human.
// Every Donna output — deltas, reports, plans, artifacts
// — lands here as a new entry by default.
// No exceptions unless Dwad says otherwise.
Karpathy's auto-research pattern is now a native Claude Code feature — /goal and /loop. The new unlock: evaluator cartridges — custom scoring algorithms that grade AI output autonomously. The combination — goal-driven agent + multi-cartridge evaluator + attempt budget — produces output that iteratively sharpens without you.
This is Entry #001's loop engineering with the inspector automated. The human graduates from checking every output to defining the scoring rubric once.
A cartridge is a self-contained scoring algorithm: output → number (1–10). Three shown:
1. Humanify Cartridge — voice, variety, predictability (AI slop detector). Only 9+ passes.
2. Hormozi AI Cartridge — scored against Alex Hormozi's public writing for marketing punch.
3. Open Rate Analytics Engine — predicted open rate from past email data + audience behavior.
Composite scoring across cartridges: newsletter required 27/30 total, no single cartridge below 8. Agent ran 5 attempts autonomously: 7.1 → 26.8 over ~40 minutes.
| Capability | Hermes Has | Source Says | Gap |
|---|---|---|---|
| Autonomous goal-driven loops | ⚠️ Manual pattern | /goal native | BIG |
| Evaluator cartridges (scoring) | ❌ None | Multi-cartridge | BIG |
| Condition-based wake-up | ⚠️ Cron (schedule) | /loop (condition) | Medium |
| Subagent swarm → single goal | ⚠️ Independent tasks | 650 experiments | Medium |
| Attempt tracking + scoring history | ⚠️ STOP AFTER only | Scoring table | Medium |
| Scoped file access for loops | ❌ Full FS | Scoped access | Small |
cronjob: "run every N minutes UNTIL condition met."Prompt engineering is dead. The new paradigm is loop engineering: AI tries → self-checks → fixes → retries → marks done. You move from being in the loop to being outside it. Four components: Goal, Checklist, Inspector, Budget.
Inner loop: Single task — try, inspect, fix, retry. AI self-corrects before delivering.
Outer loop: Recurring scheduled job running inner loops each iteration. Self-improves over time.
| Capability | Hermes Has | Loop Eng. | Gap |
|---|---|---|---|
| Scheduled recurring tasks | ✅ 17 crons | Outer loop | None |
| Self-review within task | ❌ Missing | Inner loop | BIG |
| Inspector/second pass | ❌ Missing | Grader model | BIG |
| Hard definition-of-done | ⚠️ Soft goals | Checklist | Medium |
| Skills auto-update from loops | ❌ Missing | Loop→skill | Medium |
| Per-task attempt budget | ⚠️ Repeat only | Stop after N | Small |
loop-engineer skill: GOAL + DONE WHEN + STOP AFTER → delegate → inspect → retry up to N.