Feed To Md
Analysis
The skill appears benign and purpose-aligned: it fetches a user-provided public RSS/Atom feed, converts it locally to Markdown, and only optionally writes a scoped .md file, though the feed text itself remains untrusted.
Findings (2)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
feed_url = validate_feed_url(args.url) ... xml_bytes = fetch_xml(feed_url) ... output_path.write_text(markdown, encoding="utf-8")
The script fetches a user-provided feed URL and can write generated Markdown to a file. This is the intended function and is constrained by public-URL validation and workspace-relative .md path validation, but it is still outbound network and file-write authority.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
summary = find_text(item, "content:encoded", content_ns) or find_text(item, "description") ... lines.append(summary) ... output_path.write_text(markdown, encoding="utf-8")
Text from the external feed is rendered into Markdown and may be saved for later use. That is purpose-aligned, but feed content is untrusted retrieved context.