ar-taskflow-remember

Restores task context mid-session by re-reading task-flow skill and task state files.

1|2|Updated Jul 11, 2026

Search Tags:#workflow #task management #AI assistant #claude code #session state #context recovery

Ready to run

npx skills add https://github.com/mahsanamin/agentic-repos --skill ar-taskflow-remember-mahsanamin

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ar-taskflow-remember
Source: https://github.com/mahsanamin/agentic-repos/tree/main/skills/ar-taskflow-remember
Command: npx skills add https://github.com/mahsanamin/agentic-repos --skill ar-taskflow-remember-mahsanamin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? During a long working session, an AI assistant can lose track of the current task, hallucinate the wrong phase, or suggest actions that contradict the execution plan. This Skill quickly re-orients the assistant by re-reading the authoritative task files instead of re-exploring the codebase. ## Core Features & Use Cases - Context Recovery: Re-reads the ar-taskflow workflow definition plus the task folder's execution-summary.md and execution_plan.md to rebuild accurate state. - State Summary: Reports the current task, phase, git branch, last action, and next step in a brief structured summary. - Use Case: Midway through a multi-phase implementation, the assistant starts proposing steps from the wrong phase. Say "remember" and it re-reads execution-summary.md, confirms the real current phase and branch, and continues correctly. ## Quick Start Say "ar-taskflow-remember" or "remember" when the assistant forgets the current task or drifts from the execution plan.

Frequently Asked Questions about ar-taskflow-remember

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

FAQPage Schema

How do I fix Claude forgetting context mid-session?▼

Say "ar-taskflow-remember" or "remember" to trigger context recovery. The assistant re-reads the ar-taskflow workflow and your task folder's execution-summary.md, then summarizes the current phase, branch, and next step.

What is the difference between ar-taskflow-remember and ar-taskflow-resume?▼

ar-taskflow-remember is for quick re-orientation within the same open session when the assistant drifts. ar-taskflow-resume is for full task state restoration after closing Claude and starting a new session.

Which files does context recovery read?▼

It reads the ar-taskflow skill definition first, then execution-summary.md as the source of truth for current state, and optionally execution_plan.md and prompt-understanding.md for additional context. It deliberately avoids exploring the codebase.

When should I not use quick context recovery?▼

Do not use it when starting a brand-new session after closing the assistant; use ar-taskflow-resume instead for full state restoration. It also requires an existing task folder with an execution-summary.md file to work.