Measure unknown file or command output size before reading or delegating to subagents.
npx skills add https://github.com/axiomantic/spellbook --skill smart-reading
Please help me install this Agent Skill. Skill: smart-reading Source: https://github.com/axiomantic/spellbook/tree/main/skills/smart-reading Command: npx skills add https://github.com/axiomantic/spellbook --skill smart-reading
Reading files or command output of unknown size often risks blind truncation, causing context loss and missed insights. This Skill provides a disciplined approach to capture and evaluate content without discarding important information.
Measure the size of unknown output (e.g., with wc -l) and then either read the full content or delegate analysis accordingly.
High-intent search queries and answers about installing and using this skill.
To read large log files without truncating the output, measure the file size first using commands like wc -l, then choose to either read the full content or delegate targeted summaries to subagents.
The best way to handle streaming command output of unknown size is applying a size-driven reading strategy that measures the output first, preventing blind truncation and context loss during analysis.
You prevent context loss when debugging large reports by measuring the output size before reading, which ensures explicit data integrity and avoids missing critical insights due to silent truncation.
Manage temporary artifacts generated during large output analysis by applying explicit cleanup discipline, which safely handles and removes temporary files to prevent resource leaks.
You should delegate command output reading to subagents when the initial size measurement indicates the content is too large for a direct read, allowing targeted summaries without losing full context.
Reading unknown-sized command output causes silent data loss because standard reads often apply blind truncation, discarding important information and context necessary for debugging and data integrity.
bizzybaeRun browser-based benchmarks to compare web performance across pull requests.
belokonmAnalyze media files to extract structured data and contextual insights from visual content.
AI4ScientistDiagnose and iterate on experiments with a five-step debugging workflow and structured logs.
ShenWang96Read Markdown files from allowed workspace directories with size limits.
