▸case-01 Our agent history and inbox files in the squad folder are getting huge after this long sprint. Can you run the squad context cleanup command to archive old decisions and prune logs before we start spawning new agents? | fail→pass | 8,262 | 4,555 | -45% | 1 | 1 | 0% | 1,162 | 791 | -32% | 0 | 0 | — |
▸case-02 The workspace .squad directory is accumulating too much history data from previous agent runs. Please perform a routine context hygiene sweep to compress agent logs and clear out orphaned inbox items. | fail→fail | 6,100 | 7,068 | +16% | 1 | 1 | 0% | 197 | 424 | +115% | 0 | 0 | — |
▸case-03 We are planning a heavy multi-agent fan-out session soon and need to clear out bloated history logs in our .squad/ setup. What squad utility command should we invoke to compress state and reclaim window budget? | fail→pass | 31,550 | 3,120 | -90% | 1 | 1 | 0% | 1,773 | 594 | -66% | 0 | 0 | — |
▸case-04 In our squad multi-agent setup, we want to set up automated alerts for when agent history files need context compression. Many linter defaults trigger at 100KB or 1MB, but squad sets a specific lower threshold. What specific file size limit on history.md should trigger context pruning? | fail→pass | 17,417 | 2,882 | -83% | 1 | 1 | 0% | 2,097 | 561 | -73% | 0 | 0 | — |
▸case-05 We are writing monitoring rules for our team's .squad directory footprint. Standard git repos often wait until 50MB before warning, but squad context hygiene triggers much earlier. At what total folder size threshold should context hygiene and archiving be executed? | pass→pass | 24,473 | 3,846 | -84% | 1 | 1 | 0% | 2,169 | 622 | -71% | 0 | 0 | — |
▸case-06 We have extensive archived decisions and deep history logs in .squad that need thorough deep context pruning using the CLI tool. Instead of passing custom flags like --purge or --force, which squad flag should be passed to perform a deep cleanup run? | fail→pass | 12,885 | 2,764 | -79% | 1 | 1 | 0% | 1,823 | 481 | -74% | 0 | 0 | — |
▸case-07 Before executing context compression on our squad directory, I want to simulate the cleanup to see what files and logs would be affected without altering any files. Developers often reach for --preview or --simulate. What CLI command and flag should be used? | fail→pass | 6,775 | 4,677 | -31% | 1 | 1 | 0% | 961 | 561 | -42% | 0 | 0 | — |
▸case-08 I am currently inside the squad interactive REPL session and noticed context budget running low. Users often try typing `squad nap` inside the REPL or `/clean`. What slash command should I enter in the REPL interface to prune context? | fail→pass | 11,955 | 3,212 | -73% | 1 | 1 | 0% | 1,858 | 591 | -68% | 0 | 0 | — |
▸case-09 While working inside the squad REPL, I want to perform a dry run of context compression to verify what orphaned inbox files will be removed. Rather than exiting to the shell or using `/preview`, what slash command syntax should be used in the REPL? | fail→pass | 7,475 | 2,536 | -66% | 1 | 1 | 0% | 1,276 | 515 | -60% | 0 | 0 | — |
▸case-10 During an interactive squad session, we need to execute a deep compression pass on old agent histories directly from the REPL prompt. Rather than using `/compress --full`, what slash command and option should be executed? | fail→pass | 22,906 | 2,619 | -89% | 1 | 1 | 0% | 3,802 | 491 | -87% | 0 | 0 | — |
▸case-11 We are about to launch a massive parallel workload spawning 20 sub-agents in our squad workflow. Instead of running a general git gc or waiting until agents crash, what context hygiene command should be executed right before starting this fan-out operation? | fail→pass | 14,508 | 3,812 | -74% | 1 | 1 | 0% | 2,087 | 641 | -69% | 0 | 0 | — |
▸case-12 Our squad folder contains accumulated temporary communication files in the inbox that were left behind by crashed agents. Rather than writing a custom python script or running `rm -rf .squad/inbox/*`, which built-in maintenance utility command cleans up these orphaned inbox files? | fail→pass | 8,506 | 5,276 | -38% | 1 | 1 | 0% | 1,295 | 608 | -53% | 0 | 0 | — |
▸case-13 We noticed old agent decision records cluttering active context during long sprint cycles in squad. Rather than manually moving decision files to an archive directory, what squad command archives stale decisions and compresses histories? | fail→pass | 11,168 | 3,202 | -71% | 1 | 1 | 0% | 1,673 | 600 | -64% | 0 | 0 | — |
▸case-14 When an agent session hits context window limits during a long sprint, developers often try clearing context completely with `/reset` or restarting the session from scratch. What squad utility should be run to reclaim window budget while preserving necessary state? | fail→pass | 12,863 | 3,925 | -69% | 1 | 1 | 0% | 1,870 | 742 | -60% | 0 | 0 | — |
▸case-15 We just finished a two-week team sprint with squad agents. Instead of archiving the entire workspace or manually deleting logs, what command should be included in our post-sprint workflow to clean up accumulated history and prune logs? | fail→pass | 12,365 | 3,655 | -70% | 1 | 1 | 0% | 1,896 | 688 | -64% | 0 | 0 | — |
▸case-16 Our history.md file in the squad workspace has grown to 8KB. Some team members suggest running context compression whenever history.md exceeds 5KB, but squad rules use a different limit. Does an 8KB file size meet the threshold to require a context hygiene pass? | fail→pass | 8,792 | 9,148 | +4% | 1 | 1 | 0% | 1,296 | 897 | -31% | 0 | 0 | — |
▸case-17 Checking our project status, the .squad directory total size is currently 2.4MB. A developer claims context compression is only needed once .squad reaches 10MB. Based on standard squad threshold rules, does 2.4MB warrant running the context compression tool? | pass→pass | 9,655 | 125,484 | +1200% | 1 | 1 | 0% | 1,235 | 769 | -38% | 0 | 0 | — |
▸case-18 A team member thinks squad context hygiene only deletes log files. What core context cleanup actions are actually performed when executing context hygiene in a squad workspace? | fail→pass | 11,682 | 4,577 | -61% | 1 | 1 | 0% | 1,760 | 809 | -54% | 0 | 0 | — |
▸case-19 To standardize documentation, we need to list both the terminal CLI command and the interactive REPL slash command used to preview a context compression run without making changes. Avoid using non-standard flags like --test or --preview. | fail→pass | 18,432 | 3,000 | -84% | 1 | 1 | 0% | 1,507 | 549 | -64% | 0 | 0 | — |
▸case-20 We want to initialize a new agent role named 'reviewer' in our squad setup. Should we run `squad nap` to create new agent roles, or what is the correct workflow step to define and register a new agent? | pass→pass | 11,853 | 6,013 | -49% | 1 | 1 | 0% | 1,840 | 972 | -47% | 0 | 0 | — |
▸case-21 We need to reset our git repository working directory back to HEAD and discard all uncommitted local file changes in the project. Should we run `squad nap` to revert uncommitted source code edits, or what git command should be used? | pass→pass | 7,833 | 4,518 | -42% | 1 | 1 | 0% | 1,212 | 909 | -25% | 0 | 0 | — |
▸case-22 How do we run the unit test suite for our custom agent actions in the squad workspace to verify agent code execution? Should we invoke `squad nap` or a test runner? | pass→pass | 11,580 | 31,618 | +173% | 1 | 1 | 0% | 1,668 | 741 | -56% | 0 | 0 | — |
▸case-23 We want to update the roster of human team members listed in our squad team config. Is `squad nap` the command used to modify team roster assignments? | pass→pass | 13,999 | 3,468 | -75% | 1 | 1 | 0% | 1,975 | 599 | -70% | 0 | 0 | — |