Use STOP blocks as explicit pacing primitives to manage information overload in AI-as-instructor instruction¶

Strategy · All strategies
Evidence · no claims cited

Description¶

STOP blocks are explicit pause directives embedded in module files that create hard boundaries in instruction delivery, addressing the risk that an AI agent instructed to teach a module generates thousands of words in a single response. The article states: "A STOP block serves three functions: (1) it forces the AI to pause and wait for the learner's response before continuing; (2) it provides a reflection prompt intended to consolidate learning from the preceding step; and (3) it creates a natural checkpoint where the learner can assess their understanding before moving forward." Pacing is enforced through a CLAUDE.md directive instructing the agent to halt after each STOP block, and step progress is tracked in a persistent state file that survives context compaction.

Design Implications¶

Context¶

Requirements¶

• A CLAUDE.md pacing directive and a per-session state file (CLAUDE.local.md) that is re-read at the start of each interaction

Constraints¶

• Enforcement relies on the LLM's adherence to prompt instructions rather than a hard system lock, so occasional violations are possible, though the authors report compliance to be reliable in practice

Target Learners¶

• learners receiving instruction from an AI agent in a terminal-based environment

Target Learning Goals¶

• managing extraneous cognitive load and information overload during step-by-step skill acquisition

Related Strategies¶

• Pacing
• Chunk Directions
• Segmenting

Examples¶

-

Key Sources¶

• Zain Naboulsi. (2026). Agentic Education: Using Claude Code to Teach Claude Code. arXiv. https://arxiv.org/abs/2604.17460