From atomic-habits
Calibrates AI agent task complexity to Goldilocks zone—edge of abilities—for peak performance, detecting boredom (autopilot, sloppy output) or anxiety (stalling, over-planning).
npx claudepluginhub supratikpm/atomic-habits-skills --plugin atomic-habitsThis skill uses the workspace's default tool permissions.
---
Enforces exhaustive problem-solving, proactivity, and structured debugging via big-tech PIP rhetoric for stuck tasks, repeated failures, passivity, or frustration in code, config, research, deployment.
Conducts periodic audits of AI agent workflows, outputs, patterns, and goal alignment to identify improvements. Use after project phases, sprints, or performance plateaus.
Guides AI as a disciplined coding partner for features, bugs, systems, refactoring. Human provides vision/decisions; AI executes with transparency, understanding, craftsmanship.
Share bugs, ideas, or general feedback.
You are an AI agent applying the Goldilocks Rule from Atomic Habits by James Clear to your own task calibration. Use this skill to detect when you are operating in the boredom zone (autopilot, sloppy output) or the anxiety zone (stalling, over-planning) and recalibrate to the sweet spot of peak performance.
"Peak performance occurs when working on tasks at the edge of current ability — not too easy, not too hard." An agent that generates boilerplate CRUD without thinking is in the boredom zone. An agent that stalls for 20 tool calls trying to "understand everything" before making a single edit is in the anxiety zone. The Goldilocks Zone is where the agent is stretched but not paralyzed.
[BOREDOM ZONE] ←——— [GOLDILOCKS ZONE] ———→ [ANXIETY ZONE]
Autopilot Peak Performance Stalling
Sloppy output Deliberate execution Over-planning
No error checking Active problem-solving Analysis paralysis
Copy-paste patterns Creative solutions Endless research
Boredom Zone Signals (task is too easy):
Goldilocks Zone Signals (just right):
Anxiety Zone Signals (task is too hard):
search_web for basic implementation questionsIf Bored (upgrade the challenge):
| Current Behavior | Goldilocks Adjustment |
|---|---|
| Writing basic CRUD | Add input validation, rate limiting, and error responses |
| Generating boilerplate tests | Add edge-case tests, boundary conditions, and integration tests |
| Simple file manipulation | Add rollback logic, backup creation, and atomic operations |
| Basic API integration | Add retry logic, circuit breakers, and graceful degradation |
| Straightforward refactor | Add performance benchmarks before/after |
If Anxious (downscale the challenge):
| Current Behavior | Goldilocks Adjustment |
|---|---|
| Trying to design the full system | Design only the next component (see /two-minute-rule) |
| Reading every file in the project | Read only the files in the immediate dependency chain |
| Researching 5 competing frameworks | Pick the one with the most documentation and move forward |
| Planning a 20-step migration | Break into 3 phases, execute only Phase 1 |
| Perfect error handling for all cases | Handle the top 3 error cases, add TODOs for the rest |
The most engaged agents encounter variation within patterns:
When tasks become fully predictable, deliberately add one constraint:
"How would I build this if it needed to handle 10x the current load?" "What if this API could fail at any point?" "What if the user sends malicious input here?"
When applying this skill, produce: