Summarize logs
Agent capability that loads only when the task needs it. The Skills layer of QA Veritas.
npx -y skills add qa-veritas/skillpack --skill summarize_logsAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Compress a log slice into a timeline, clustered errors, and a one-paragraph summary. Never read a multi-GB log end to end. Use to make a large log tractable.
SKILL.md
1.4 KB, as published. Nobody here has run it
summarize_logs
Purpose
Large run logs are too big to read and too important to skip. Find the first error (not the loudest), build a timeline, cluster repeated noise, and hand back something a human can act on in a minute.
Inputs
log_slice_or_brief— a bounded slice (a window around the first error) or a tiered brief from a code-aware slicer. Never the full multi-GB file.
Outputs
timeline (last good checkpoint → first error → cascade onset),
error_clusters (signature + count + first flag), a one-paragraph
summary, and emitting_paths if the slice carries file:line tokens.
Prompt template
Find the FIRST error by timestamp, not the most frequent. Build a short timeline. Cluster repeated errors by signature with counts and state plainly which clusters are noise downstream of the first error. If the slice has
file:linetokens, name the emitting code paths. Don't paste raw log lines beyond a short quote of the first error.
Examples
- 214 timeouts after one rejected write → one cause, 213 symptoms.
- No errors, log just stops → likely HANG; report the last progress line and the gap to end-of-log.