dashboard_url: https://dashboard.kismetpets.com/context/health/lint/1/ generated_at: ‘2026-04-28T18:31:58.992741+00:00’ id: 1 scope_id: null scope_type: ”

Lint Report — 2026-04-28 14:31

Scope: global

Context Lint Report — 2026-04-28 (global)

Score: 70/100

Orphan results (119)

The compounding-memory layer just shipped — every existing skill result that landed before Slice 2 has no ResultLink to an Entity or Concept yet. This is expected on day-one and is the biggest lever for raising the score.

Distribution by skill:

  • positioning: 44
  • dtc_ads: 30
  • market_research: 27
  • direct_response_copy: 7
  • voice_mining: 5
  • front_end_design: 2
  • measurement: 2
  • orchestrator: 2

Suggested action: run /context-ingest-result over the backlog. Within a single steward run that’s bounded to ~10 ingests; the rest will roll forward into the next run.

Stale entities (0)

None — no entities exist yet, so nothing can be stale. This number will become meaningful after the first ingests land.

Duplicate candidates (0)

None — same reason as stale: no entities yet. The duplicate detector will start surfacing candidates once /context-ingest-result runs and the Entity table fills up.

Briefs needing compaction (0)

None — no ContextBrief rows exist yet. The first deltas will get appended by /context-ingest-result (it seeds new entities into the right scope brief) and from regular skill runs once the preamble loads briefs into context.

  1. Run /context-ingest-result — top priority, every ingest moves an orphan into the synthesis layer and seeds entity/concept pages.
  2. After ingests are flowing, re-run /context-lint weekly to catch the first stale entities and duplicate candidates.
  3. Defer /context-merge-duplicates and /context-compact-brief until the layer has actual content to work over (they’ll no-op today).