instruction-adherence

What problem does it solve?

This Skill ensures AI agents reliably follow instructions, especially during multi-step tasks, mid-task detours, or when frequent re-anchoring to the main objective is necessary.

Core Features & Use Cases

• Instruction Snapshot: Captures the core objective, mandatory actions, and constraints.
• Task Ledger: Structures tasks with clear parent/child relationships to prevent detours from derailing the main goal.
• Re-anchor Loop: Periodically verifies alignment with the original objective after subtasks or tool calls.
• Detour Protocol: Manages intermediate tasks, ensuring a clean return to the main objective.
• Use Case: When an agent is asked to "write a report and also improve the CLI tooling," this Skill helps it complete the report first, then manage the tooling improvement as a distinct sub-task before returning to any further report-related steps.

Quick Start

Use the instruction-adherence skill to capture an instruction snapshot and build a task ledger for the current objective.