Manage named file-based scratchpads for recording intermediate findings.
npx skills add https://github.com/misunders2d/ori --skill scratchpad-skill
Please help me install this Agent Skill. Skill: scratchpad-skill Source: https://github.com/misunders2d/ori/tree/main/skills/scratchpad-skill Command: npx skills add https://github.com/misunders2d/ori --skill scratchpad-skill
This Skill provides a file-based scratchpad outside your conversation context to record intermediate findings, large outputs, or evolving data across multi-step tasks.
Create a named scratchpad, use scratchpad_write after each tool output to record findings, and review the full history with scratchpad_read when ready.
High-intent search queries and answers about installing and using this skill.
To record intermediate findings outside the current conversation context, use a file-based scratchpad to write and manage multi-step insights, preventing prompt bloat during long workflows. Named pads track evolving data specifically.
The best way to manage large tool outputs during multi-step research is writing them to a file-based scratchpad. This keeps evolving data outside your conversation context, allowing you to review full history later without bloating prompts.
Yes, you can use a persistent scratchpad for iterative problem solving across long workflows. It provides file-based memory outside your conversation to track progress, results, and decisions by creating named pads for distinct tasks.
You organize and manage named scratchpads for task-specific notes using available file operations. You can write, read, replace, clear, and list named pads, allowing structured tracking of intermediate findings and iterative collaboration data.
A file-based scratchpad helps with memory limitations in multi-step research by storing intermediate findings and large outputs externally. This prevents conversation context bloat, ensuring you maintain efficient prompts while tracking iterative progress.
You should avoid storing intermediate findings in the main conversation context during large multi-step workflows. Storing large tool outputs directly causes prompt bloat, so using an external file-based scratchpad keeps your main context streamlined.
ArogyaReddyDefine and implement SLIs and SLOs with error budgets and alerting.
matt-fell-aiSwarm intelligence for multi-agent orchestration and problem-solving.
scawfulDiscover AFS context roots, validate scaffolding, and read or write scratchpad memory.
bykoleksii-hardoManage persistent agent context using filesystem-based storage techniques.
