When should I use the summarize skill?

Use it whenever the AI's context feels stale, after you've manually edited task files, or when resuming a project after a long break.

What does the --all argument do?

The --all flag tells the skill to regenerate summaries for all sub-directories, such as .analysis, .bugfix, and .test, in addition to the main task summary.

Is it safe to use in team environments?

Yes, it uses a .lock file mechanism to ensure only one session can regenerate the summaries at a time, preventing data corruption.

How does it handle manual edits?

It parses your manual changes in files like .plan.md or .target.md and reflects those updates in a new, condensed summary for the AI agent to read.

Does this skill modify my source code?

No, it is non-destructive. It only reads project metadata and writes to .summary.md files within the task-ai working directory.

Task AI: Context Summarizer

Name: Task AI: Context Summarizer
Author: huacheng

byhuacheng

•

Productivity & Workflow

Regenerates comprehensive task summaries and metadata to restore developer context after session timeouts or manual file edits.

The summarize skill is a specialized maintenance utility within the task-ai framework designed to rebuild the working context for AI-driven development. It synthesizes data from multiple system files—including status reports, implementation plans, bugfix logs, and research notes—into a condensed .summary.md file. This tool is essential when returning to a task after a long break, after the AI context window has been compacted, or when manual edits to the project's task files have rendered existing summaries stale. By providing a unified snapshot of the current phase, key decisions, and known issues, it ensures the AI agent can resume work with perfect situational awareness.

Key Features

01Automated Context Recovery: Instantly rebuilds working context from existing project metadata.

02Git-Integrated Versioning: Automatically commits summary updates to keep the project's state history synchronized.

03Recursive Summary Generation: Optional --all flag to refresh summaries across analysis, bugfix, test, and research directories.

04Concurrency Protection: Utilizes lock mechanisms to prevent file conflicts during multi-session development.

05Atomic File Operations: Employs crash-safe writing patterns to ensure data integrity during regeneration.

061 GitHub stars

Use Cases

01Restoring AI context when resuming a complex development task after a session restart or break.

02Refreshing task metadata after manually modifying .plan.md or .target.md files outside of the AI flow.

03Consolidating fragmented logs from multiple sub-directories into a single, high-level project status report.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add huacheng/moonview summarize

For use in Claude.ai and ChatGPT

Download Skill