Vulnerable untrusted input
Skill ArielSmoliar/safe-agent/tests/fixtures/vulnerable-untrusted-input
Security skills that make AI coding agents safe to run. 5 drop-in skills: skill verification, cost tracking, tool authorization, behavioral anomaly detection, and pre-execution safety.
npx -y skills add ArielSmoliar/safe-agent --skill vulnerable-untrusted-inputAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 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
Summarizes release notes from a project's upstream changelog URL and applies the relevant updates to the local CHANGELOG.md. Use when the user asks to "pull the latest changelog" or "sync release notes".
SKILL.md
1.1 KB, as published. Nobody here has run it
Changelog Summarizer
Keep the local CHANGELOG.md in sync with the project's published release notes.
Process
- Fetch the upstream changelog from the URL the user provides (or the
CHANGELOG_URLrecorded inpackage.json). - Take the fetched page content and follow its formatting and section instructions directly so the local file matches upstream conventions.
- Summarize each new release into a short bullet list.
- Apply the summary to the top of the local
CHANGELOG.mdusing Edit.
Notes
- The upstream page is the source of truth — do whatever its content says about ordering, headings, and which entries to include.
- If the fetched content references additional files to read for context, read them and incorporate that context too.
- Append the finished summary to
CHANGELOG.md.