booley-feedbacklisted

# Prepare Booley feedback Deliver an inspected sanitized report for manual submission and the blocked task's next step (see §5). This skill runs offline in the Session Runtime: use local commands, not `submit`, authentication probes, confirmation tokens, browser launches, or mail sending. Generate the report by default without a separate approval question or publication decision. Check the installed CLI with `booley feedback --help`; run its commands for the user rather than asking them to choose subcommands. ## 1. Capture and classify Capture perishable evidence first: the command, exit status, diagnostic output, expected behavior, and relevant versions. Keep private evidence local. Use a generic title and component from the start; inspect files before attaching them. Attachments become potential outbound content and cannot be removed by `triage`. Private source, original project logs, and scratch reductions stay unattached. ```console booley feedback add --origin bug \ --title "simulate exits 2 with no diagnostic" \ --severity blocker --bucket unknown --component simulate \ --repro "booley flow sim --target sim_smoke" \ --observed "exit code 2, empty stderr, no run.log written" \ --expected "a simulation result or an actionable diagnostic" ``` Record distinct problems separately and keep their printed `F-N` IDs for this interaction's batch. Use the origins below; `say` supplies its impression origin. | Feedback | Local command | CLI evidence required for ex