← Autoresearch Loop · Prompt Deck · Pictures

Story Contract​

Stories are test contracts. Tests must be RED before implementation starts. GREEN = value delivered.

S1 — PRD metric enforced at creation​

Trigger: Agent creates a PRD and defines the Intent Contract outcome

Checklist:

• Metric-definer enforces a queryable formula (not prose) with named data source, threshold, and unit
• PRD without a formula is blocked — not silently accepted
• PRD without a named data source fails the gate

Forbidden: Prose metric accepted without formula. Metric with no data source passes gate.

Evidence: integration — tests/skills/metric-definer.spec.ts

Commission Result: ⬜ PASS / ⬜ FAIL Notes: (findings)

S2 — Validation outcome propagates to frontmatter​

Trigger: Engineering validates a PRD (RED → GREEN) and outcome-validator posts results

Checklist:

• proof-to-story writes actual_metric_value, validation_outcome, and pain_reduction_delta to PRD frontmatter
• Fields are machine-readable — not buried in spec prose
• Next score-prds run reads the new frontmatter fields

Forbidden: Validation results stored only in spec prose. No frontmatter update.

Evidence: integration — tests/skills/proof-return-signal.spec.ts

Commission Result: ⬜ PASS / ⬜ FAIL Notes: (findings)

S3 — Validated PRDs score higher​

Trigger: Score-prds runs on a PRD with validation_outcome in frontmatter

Checklist:

• Validated PRDs receive a confidence boost in priority score
• Failed PRDs flagged for re-evaluation — not silently kept at current score
• Unvalidated PRDs score differently from validated PRDs

Forbidden: Validated and unvalidated PRDs scored identically. Evidence quality ignored.

Evidence: unit — tests/skills/score-validated.spec.ts

Commission Result: ⬜ PASS / ⬜ FAIL Notes: (findings)

S4 — Session experiment logged automatically​

Trigger: Agent session completes (overnight or manual)

Checklist:

• Session-experiment-logger reads Comms, aggregates deltas, writes structured log
• Morning report is diff-readable — not scattered across channel messages
• Log produced for every session, not only sessions with changes

Forbidden: Session results scattered across Comms with no aggregation. No morning report.

Evidence: integration — tests/skills/session-logger.spec.ts

Commission Result: ⬜ PASS / ⬜ FAIL Notes: (findings)

S5 — Full loop runs from one trigger​

Trigger: Operator triggers "run loop" on top uncommissioned PRDs

Checklist:

• Loop-orchestrator chains scaffold → activate → validate → measure → story for each PRD
• Loop stops at budget cap — does not continue past limit
• Loop halts on metric regression — does not continue to next PRD
• Metric check runs before each iteration — not skipped

Forbidden: Loop runs without metric check. Loop continues past budget. Loop ignores regressions.

Evidence: e2e — tests/skills/loop-orchestrator.spec.ts

Commission Result: ⬜ PASS / ⬜ FAIL Notes: (findings)

Build Contract​

Metric Enforcement​

Return Signal​

Experiment Infrastructure​

Principles​

SIO:

• Situation: When the priority stack has scored PRDs with Red tests waiting, and an agent has overnight compute budget available.
• Intention: The pain-to-proof cycle should run autonomously — measure, iterate, validate, propagate results — without seven separate human triggers.
• Obstacle: Metrics aren't queryable (prose, not formulas). Validated outcomes don't propagate back to frontmatter. No conductor chains the skills.
• Hardest Thing: Making the loop trustworthy enough to run unsupervised. A bad loop that produces false greens is worse than no loop.

Why Now: The skill chain exists. The CLIs exist. Karpathy proved the pattern works at research scale. The gap is three wires, not a new architecture.

Design Constraints: Activation over creation. Wire existing skills before building new ones. Every metric must name its data source. Every loop iteration must log its experiment.

Performance​

• Priority Score: 1200 (Prepare)
• Quality Targets: prd_full_cycles_per_month >= 1. North Star queryable for all Active PRDs within 30 days.
• Kill Signal: Loop runs for 90 days with zero PRDs completing a full cycle. Or: metrics are defined but never queried.

Platform​

• Current State: 7 skills built, 0 wired into a loop. 7/7 Active PRDs have Queryable: No.
• Build Ratio: 30% new (3 skills + 1 CLI), 70% wiring (3 skill upgrades + integration).

Protocols​

• Build Order: Phase 1 (metric-definer + measure-north-star.ts) → Phase 2 (proof-to-story upgrade + score-prds upgrade) → Phase 3 (session-logger + loop-orchestrator)
• Commissioning: L2 when metric-definer enforces on one PRD. L3 when one PRD completes full cycle. L4 when prd_full_cycles_per_month >= 1 sustained for 60 days.
• First Target: Sales CRM (L2-L3, clearest metric: won_deals / total_deals > 0.30, real data in Postgres)

Players​

• Demand-Side Jobs:
  • As a platform operator, I want overnight agent sessions to produce structured experiment logs so I wake to a readable diff instead of scattered Comms.
  • As a PRD author, I want validated outcomes to automatically boost the next scoring cycle so proven demand compounds without manual re-scoring.
  • As an agent, I want to measure a PRD's North Star via CLI so I can determine whether my iteration improved or regressed the metric.

Context​

• PRD Index — Autoresearch Loop
• Prompt Deck — 5-card pitch
• Pictures — Pre-flight maps