Introduction to Subagents

• QuestionWhat is "Two valid spawn triggers"?Tap to revealAnswerSpawn a subagent only for parallel work or to keep messy intermediate context out of the parent. Cost, window size, and tool count are NOT reasons.
• QuestionIn "Two valid spawn triggers", explain:Tap to revealAnswerTwo and only two triggers: parallelism, isolation.
• QuestionIn "Two valid spawn triggers", explain:Tap to revealAnswerParallelism: independent work, fan-out reduces wall-clock latency.
• QuestionIn "Two valid spawn triggers", explain:Tap to revealAnswerIsolation: intermediate context kept out of parent.
• QuestionIn "Two valid spawn triggers", explain:Tap to revealAnswerNon-triggers: cost, window size, tool count, 'cleaner' — common distractors.
• QuestionAnalogy for "Two valid spawn triggers":Tap to revealAnswerLike delegating to a teammate at work: do it when you can run two efforts at once or when you don't want their messy whiteboard in your office. Not 'because it feels organised.'
• QuestionWhat is "Critique underbriefed subagent prompt"?Tap to revealAnswerSubagents have no memory of the parent conversation. Write the brief as if explaining to a teammate who just walked in: goal, context, constraints, expected output.
• QuestionIn "Critique underbriefed subagent prompt", explain:Tap to revealAnswerSubagents have no parent-conversation memory. Brief is all they have.
• QuestionIn "Critique underbriefed subagent prompt", explain:Tap to revealAnswerBrief = goal + relevant context + constraints + expected output shape.
• QuestionIn "Critique underbriefed subagent prompt", explain:Tap to revealAnswerSmell phrases: 'based on the above', 'as discussed', 'continue from where we left off'.
• QuestionIn "Critique underbriefed subagent prompt", explain:Tap to revealAnswerFix: write as if briefing a smart colleague who just walked in.
• QuestionAnalogy for "Critique underbriefed subagent prompt":Tap to revealAnswerIt's like emailing a contractor. 'Finish the kitchen as discussed' doesn't work if they weren't in the discussion. 'Install the granite counter from order #4421, white grout, by Friday, leave the sink for the plumber' works.
• QuestionWhat is "Cost of unnecessary subagent splits"?Tap to revealAnswerSubagents that don't run in parallel and don't need isolation are pure cost — they pay setup, round-trip, and integration overhead with no benefit.
• QuestionIn "Cost of unnecessary subagent splits", explain:Tap to revealAnswerCost = N× tokens (setup, round-trip, integration) for N unnecessary subagents.
• QuestionIn "Cost of unnecessary subagent splits", explain:Tap to revealAnswerNo latency win and no isolation win = pure waste.
• QuestionIn "Cost of unnecessary subagent splits", explain:Tap to revealAnswerSequential dependent work should stay inline.
• QuestionIn "Cost of unnecessary subagent splits", explain:Tap to revealAnswerDetection: parent serialises subagents (waits between briefs) — worst of both worlds.
• QuestionAnalogy for "Cost of unnecessary subagent splits":Tap to revealAnswerLike hiring three different contractors to do three sequential parts of a small job. You pay each one's setup time and travel — and they still wait for each other. One contractor doing all three would be cheaper and the same speed.
• QuestionWhat is "What parent does NOT see from subagent"?Tap to revealAnswerThe parent only sees the subagent's final summary. The tool calls, intermediate outputs, and dead-ends stay in the subagent's context.
• QuestionIn "What parent does NOT see from subagent", explain:Tap to revealAnswerContext isolation is unidirectional. Parent → brief in. Subagent → summary out.
• QuestionIn "What parent does NOT see from subagent", explain:Tap to revealAnswerParent does NOT see: tool-call trail, intermediate outputs, dead-ends, raw reads, failed attempts.
• QuestionIn "What parent does NOT see from subagent", explain:Tap to revealAnswerParent sees only what the summary contains.
• QuestionIn "What parent does NOT see from subagent", explain:Tap to revealAnswerDesign implication: summary must contain everything the parent needs to act on.
• QuestionAnalogy for "What parent does NOT see from subagent":Tap to revealAnswerIt's like a consultant report. You get the recommendations and a brief explanation. You don't get every email, draft, or whiteboard photo from inside their team.

1. 1Diagram the core idea

   Analyse~10 min

   In your own words, diagram the central idea of "Two valid spawn triggers". Aim for one paragraph, no jargon you haven't earned.

   Done when: A non-expert could read your paragraph and explain it back to you.

   Re-read the source concept

2. 2Solve it to a real situation

   Apply~15 min

   Find a real situation from your own work / life where "What parent does NOT see from subagent" applies. Solve the idea — describe the situation, the move you'd make, and the alternative you considered first.

   Done when: You can name a specific situation (not a hypothetical), state your move, and explain the trade-off in one paragraph.

   Re-read the source concept

3. 3Teach it back

   Evaluate~5 min

   Pick someone (a colleague, a peer, a rubber duck) and explain the section to them in under 5 minutes. Use your own example, not the one from the lesson.

   Done when: Your listener can repeat the core idea + the trade-off back to you without prompting.

Why this matters. Reading + quizzing prove recall; applying it in a real situation is the only evidence of transfer. Treat at least one of these as non-negotiable before you move on.