self-improving-prompt

Good prompt refinement reduces ambiguity and execution risk. It should not add routine friction to already clear work.

Transform vague or high-risk user prompts into clearer, actionable, verifiable versions. Use this skill selectively: only when refinement materially improves scope, constraints, output expectations, or acceptance criteria.

This skill works standalone. When paired with self-improving-session, the two skills form a closed loop:

• self-improving-prompt shapes how the current task is framed and executed
• self-improving-session learns durable workflow rules from the behavior patterns that emerge during those sessions

self-improving-session should learn from behavior and outcomes, not from storing the refined prompt text itself.

Security and Runtime

• Instruction-only skill; no bundled scripts or external services required
• No credentials, API keys, or network access required for normal use
• No persistent configuration changes are required to use the skill
• Optional automation snippets belong in README only and are not part of the core runtime behavior