context-summarizer

What problem does it solve?

Long-running AI software development sessions frequently hit context token limits, causing loss of critical task progress, key technical decisions, and current work status when context overflows, requiring manual rework to resume work.

Core Features & Use Cases

• Automated Triggered Compression: Automatically compresses context when reaching 25 messages, 30% token usage, 60% forced threshold, or detecting hallucination loops, no manual prompting required.
• Critical Artifact Preservation: Retains completed work, key technical decisions, current task status, pending tasks, and identified risks while discarding redundant intermediate outputs.
• Seamless Session Resumption: Automatically loads previous context summaries when starting a new session, eliminating the need for manual context pasting. Use case: For AI-assisted software development teams working on long-horizon projects, this skill ensures no progress is lost during context overflow and allows seamless work resumption across sessions without re-explaining prior work.

Quick Start

Invoke the context-summarizer skill to automatically compress the current task context, preserve key decisions and progress, and resume prior work seamlessly when context usage approaches the token limit.