Starlight Intelligence
Explore

Find your way through Starlight

Follow a completed mission, give your next project a brief, or build on the foundation.

Follow the workMissionsMake somethingMemory StudioMake it your ownFor buildersBuild togetherCommunity
Build once. Bring it with you.Capability FoundryFollow the thinkingResearchSee what is releasedOpen sourceQuestions and their evidenceStarlight AtlasJudgment that stays yoursTrust playbook2 to 144 perspectivesPlan a teamThe project and founderAbout StarlightYour first useful workflowStart with your agentTurn an idea into a crewMission composerMeet the crew and every seatAgentsCheck the prerequisitesInstall routerInspect the packagesDepartmentsPractice firstLearnGive an agent a jobAgent KitsSee what each agent contributedFollow a missionPick a working setTeamsSee the visual worldImagesFit to your workConfigureShared planePartnershipsA job in motionWorkflowsWho talks to whomAgent graphKeep the keyFoundationOne buildWorkshopsFour movesPathIdeas and evidenceTechnical notesContractsDevelopers

LAB 01 · CONTRIBUTOR DIAGNOSTICS

Same task.
Different ways to work.

What changes when people work alone, with AI, or through an agent team? Make the work inspectable. Keep the judgment yours.

Prototype · this tab only. Use synthetic data and participant codes. Export to keep your observations. Shared sessions and authenticated group access are not enabled here.

01 / THE BRIEF

Leave the next contributor a working handoff.

Produce a diagnostic runner, workflow and README from the supplied repository. Select existing test and lint scripts. Exclude deployment. A reviewer must be able to trace every check to the source.

01

human only

People inspect and prepare the handoff; no generative AI.

35 minutes · same source · same rubric
02

ai assisted

People direct each AI interaction and edit the result.

35 minutes · same source · same rubric
03

delegated agent

People set the brief and review the result; the existing mission runtime coordinates four roles.

35 minutes · same source · same rubric

02 / FOLLOW THE HANDOFF

Inspect what each role contributes.

This rehearsal runs the existing Starlight mission engine with scripted outputs. It makes no model calls and provides no evidence of AI performance.

  1. investigatepending
  2. designpending
  3. buildpending
  4. evaluatepending
Open the mission workbench ↗

03 / RECORD THE EVIDENCE

Quality first. Then time and cost.

Scores: 0 absent, 1 materially flawed, 2 needs revision, 3 acceptable, 4 strong. Acceptance requires 75/100, correctness and evidence at least 3, and all critical checks passed. Observations are self-reported, not independently verified.

Session observations · no scientific inference
Condition / evidenceQualityAcceptanceCostHuman / elapsed
No observations yet.

04 / CARRY IT FORWARD

Turn one useful result into a repeatable practice.

Export the evidence. Reopen the work in Missions. Agree on a named owner, one recurring task and a 14-day adoption check before adding more autonomy.

Plan a team experiment ↗