Skip to main content

← Autoresearch Loop · Prompt Deck · Spec

Pre-Flight Maps

Classification: Platform. Required maps: Outcome, Dependency.

Outcome Map

What observable state changes prove this PRD succeeded?

OutcomeMeasureBaselineTargetTimeline
North Stars queryableCount of Active PRDs with Queryable: Yes0/77/730 days
Full cycles completedPRDs completing pain → proof → marketing in one month0≥1/month60 days
Return signal wiredPRDs with actual_metric_value in frontmatter0≥145 days
Morning reports generatedSessions producing structured experiment logs0≥1/week30 days

Dependency Map

What must exist before each feature can work?

metric-definer ──────────────────────────────┐
│ │
▼ ▼
measure-north-star.ts engineering-handoff Gate 8
│ (fails if Queryable: No)

outcome-validator (reads scalar)


proof-to-story upgrade ──► score-prds upgrade
│ │
▼ ▼
session-experiment-logger (proven demand compounds)


loop-orchestrator (chains all of the above)
FeatureDepends OnBlocking
LOOP-001 metric-definerNothingLOOP-002, Gate 8 upgrade
LOOP-002 measure-north-star.tsLOOP-001 metric definitionsLOOP-005
LOOP-003 proof-to-story upgradeNothingscore-prds upgrade
LOOP-004 session-experiment-loggerComms CLILOOP-005
LOOP-005 loop-orchestratorLOOP-001, LOOP-002, LOOP-003, LOOP-004First loop execution

Context

Questions

What's the minimum viable loop — the smallest set of wires that produces one full cycle?

  • Can LOOP-001 + LOOP-003 deliver value without the full orchestrator (LOOP-005)?
  • Is the Sales CRM metric (won_deals / total_deals > 0.30) the right first target, or should we start with a simpler metric?