Ai Product Monitor
PassAudited by ClawScan on May 6, 2026.
Overview
No malicious behavior is evident; the skill appears to perform the advertised RSS, search, screenshot, and reporting workflow, with some disclosed web activity and minor metadata/dependency gaps.
Install/run only if you are comfortable with it fetching configured RSS feeds, sending product-launch titles to web search, visiting launch URLs to capture screenshots, and storing results under the skill’s `data/` directory. Review `references/feeds.yaml` first and consider running stages individually for more control.
Findings (3)
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.
When you run the pipeline, it may search the web and open product URLs for screenshots based on RSS content.
The script uses local OpenClaw search and browser screenshot tools with RSS-derived titles and URLs. This matches the skill purpose and does not use shell interpolation, but it means running the skill can send queries to web search and visit product pages automatically.
subprocess.run(["openclaw", "web-search", query, "--count", "3", "--format", "json"], ...); subprocess.run(["openclaw", "browser", "screenshot", "--url", url, "--out", str(out_file)], ...)
Review or customize `references/feeds.yaml` before running, and run individual stages if you want more control over which URLs are searched or captured.
The skill may not run as expected unless Python and the OpenClaw CLI tools are available, even though the registry metadata does not declare them.
The registry requirements under-declare the runtime expectation shown in the skill documentation. The visible code also uses optional `openclaw` CLI helpers for search and screenshots.
metadata: "Required binaries ... none"; SKILL.md: "requires_bins: [python3]" and "python3 scripts/pipeline.py"
Confirm the local runtime requirements before use; publishers should align registry metadata with SKILL.md and code behavior.
Reports may reflect whatever appears in the configured feeds or search results, including stale, promotional, or misleading content.
The workflow persists public RSS and search-derived context, then reuses it in later stages for trend reporting. This is intended behavior, but the source content is external and could be inaccurate or manipulated.
RSS Feeds → raw_launches.json → enriched_launches.json → screenshots/ → trend_report.md
Treat generated trend reports as draft analysis and review the underlying feed items before relying on conclusions.