It could be usefull to have bug-reporting skill,
which could be utilized when/after analyzing issue with claude (be it test failure or even env setup issue),
to assist with opening new bug report.
Especially I would want to use it in active analysis session once the triage outcome seems appropriate (btw this may also point at neeed for triage skill).
Few ideas of what it could help with:
- ensure correct ticket destination (jira, potentially github/... too)
- ensure correct project or component (based on the analysis outcomes it may already know if issue is more likely in helm, operator, ansible or trustify itself)
- add summary/description of the issue
- add technical analysis
- add evidence/logs used to arrive at the outcome
- sanitize them for any specific IPs, hostname or credentials (may be better left for dedicated tool or in addition to it?)
- attach any needed labels, affected-versions
- it may be specifically NOT adding solution into the description (leave that on separate/dedicated task, although in simpler cases it can be quite correct already due to access to tooling mcp's and such) ... or maybe add suggested solution only as proposal in comment?
It could be usefull to have bug-reporting skill,
which could be utilized when/after analyzing issue with claude (be it test failure or even env setup issue),
to assist with opening new bug report.
Especially I would want to use it in active analysis session once the triage outcome seems appropriate (btw this may also point at neeed for triage skill).
Few ideas of what it could help with: