All static findings are documentation artifacts in SKILL.md, including command examples, inline option names, example.com URLs, and sibling markdown links. No prompt injection, data exfiltration intent, or executable code was found in the reviewed file.
All static findings are documentation artifacts in SKILL.md, including command examples, inline option names, example.com URLs, and sibling markdown links. No prompt injection, data exfiltration intent, or executable code was found in the reviewed file.
The static findings are mostly false positives from Markdown code fences, code spans, example URLs, and relative documentation links. The real risk is operational: this skill authorizes a networked Firecrawl CLI to fetch user-provided URLs and optionally write output files, so users should review URLs and output paths before running it.
The skill permits Bash execution of firecrawl and npx firecrawl commands. This is expected for a scraping skill, but it can fetch arbitrary user-provided URLs through an external service.
The allowed-tools block and command examples directly authorize external CLI execution. The purpose is legitimate, but the commands perform network requests against user-provided URLs.
The command examples and option table allow an output path with -o or --output. The examples use .firecrawl paths, but a careless path could overwrite local files.
The file output option is clearly documented, but the examples steer users toward a local .firecrawl directory. The risk depends on the path chosen at runtime.
Remote Query Mode May Send Page Questions Externally
The --query option asks a question about page content and costs extra Firecrawl credits. This may send the prompt and retrieved page content to the service.
The query feature is explicitly documented, but SKILL.md does not describe its transport or data handling details. The privacy concern is plausible because the skill uses a remote scraping CLI.
Static false positives ignored (4)
These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.
Markdown Backticks Misclassified as Ruby Shell Execution
The reported backtick findings occur inside Markdown code fences or inline command references. They document CLI usage and are not Ruby code execution in this skill file.
The file is Markdown, and the matched backticks delimit examples or option names. No Ruby source file or executable Ruby expression is present.
The reported weak cryptography locations are front matter text and a Markdown table header. No cryptographic algorithm or hashing operation appears in the file.
The cited lines contain description text and option table formatting. I found no evidence of MD5, SHA1, DES, or another weak cryptographic operation.
Static analysis flagged 30 patterns that are all false positives. External command findings are documentation examples for the Firecrawl CLI with properly restricted bash patterns (firecrawl *). Network URLs are example placeholders in documentation, not hardcoded endpoints. Filesystem references are markdown cross-links, not path traversal. The skill legitimately wraps the Firecrawl CLI for web scraping with appropriate command restrictions.
1
Files scanned
69
Lines analyzed
4
Review items
0
False positives ignored
Capability review items (1)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
The skill executes firecrawl CLI commands via bash. Commands are restricted to 'firecrawl *' and 'npx firecrawl *' patterns, limiting injection risk. URL and option inputs could potentially be crafted to escape intended command structure.