goal-anchor

Inject OBJECTIVE blocks into prompts and enforce drift checks every five turns.

15|6|Updated May 7, 2026

Search Tags:#multi-turn #agent-prompt #phase-transition #prompt-injection-prevention #drift-prevention #goal-anchor

Ready to run

npx skills add https://github.com/MinhThang1009/dotclaude --skill goal-anchor

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: goal-anchor
Source: https://github.com/MinhThang1009/dotclaude/tree/main/plugins/subagent-system/skills/goal-anchor
Command: npx skills add https://github.com/MinhThang1009/dotclaude --skill goal-anchor

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents goal drift in long-running or multi-phase AI agent workflows by ensuring every prompt starts with a clearly defined OBJECTIVE.

Core Features & Use Cases

Injects an OBJECTIVE block at the start of the subagent prompt to anchor the intended goal.
Enforces drift checks every 5 turns and provides a phase-transition protocol with compaction to preserve objective integrity.
Supports different thresholds for multi-agent reviews and general tasks, and includes a structured, repeatable process for resuming after interruptions.

Quick Start

Inject the OBJECTIVE block at the start of the subagent prompt before spawning a multi-turn agent.

Frequently Asked Questions about goal-anchor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I prevent AI agent drift during long-running multi-turn workflows?▼

To prevent agent drift, inject an explicit OBJECTIVE block at the start of the subagent prompt before spawning multi-turn tasks. This anchors the intended goal and enforces periodic drift checks every five turns to maintain objective integrity.

What is the best way to preserve agent objectives after a session interruption or compaction?▼

Preserve agent objectives after a session interruption by using a structured phase-transition protocol. This protocol applies context compaction and re-injects the OBJECTIVE and SCOPE blocks into the prompt to cleanly resume the task without losing the original goal.

When do I need to anchor objectives for multi-phase agent tasks?▼

Anchor objectives for multi-phase agent tasks when workflows exceed 15 turns, when chaining phases with accumulated context, or when resuming after a session interruption. This prevents the agent from losing sight of the original goal over extended operations.

How do I chain multi-phase agent tasks without losing the original goal context?▼

Chain multi-phase agent tasks without losing context by applying a clean phase-transition protocol that includes context compaction. This protocol systematically clears accumulated drift and re-injects the OBJECTIVE block before starting the next phase.

Does multi-agent prompt injection prevention require different drift check thresholds?▼

Multi-agent prompt injection prevention supports different drift check thresholds for multi-agent reviews compared to general tasks. This flexibility ensures objective anchoring remains effective across varying complexities and agent interaction levels.