| name | file-summarize |
| description | Produce a short Markdown summary of a text file (line count, top tokens, first/last lines). Use when the user asks "what's in this file" or wants a quick characterisation. |
| tags | ["text","summary","exploration"] |
File summariser
Use this skill when the user wants a quick characterisation of a text
file but does NOT want full content reproduction.
Steps
- Use
read_text(path=<file>) to load the file.
- Count: total lines, total chars, distinct words (split on
whitespace, lowercased, stripped of
.,;:!?"'()).
- Identify the 10 most-frequent words, ignoring this stoplist:
the a an of to in for and or is are was were be been being it its this that these those on at by from as with.
- Capture the first 3 non-empty lines and the last 3 non-empty
lines.
- Write a Markdown report to
<stem>.summary.md with sections:
## Counts, ## Top words, ## First lines, ## Last lines.
Output contract
The user expects <stem>.summary.md. Report the path in
done(answer=...).