Anti-SEO Researcher
v1.0.2Anti-SEO deep consumer research tool. When a user wants to buy a product or make a consumer decision, use this Skill. Automatically detects user language and...
⭐ 0· 176·0 current·0 all-time
byAmos@amoshc
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
The name/description match the included scripts and workflow: multi-source search, credibility scoring, conflict resolution, and report generation. The code implements search fallback scraping, scoring, and report generation which are coherent with an 'Anti-SEO Researcher'. There are no environment variables, cloud creds, or unrelated binaries requested.
Instruction Scope
SKILL.md confines actions to searching (web_search or a fallback scraper), fetching pages (web_fetch), running bundled Python scripts (execute_command), and reading/writing local files (cache and output reports). It does not instruct reading arbitrary host system files or contacting hidden/external endpoints beyond normal web fetching and search scraping. However, it does require invoking bundled Python scripts via shell which will execute arbitrary Python code present in the skill directory — the host agent will run those scripts and they write caches and output files.
Install Mechanism
There is no install spec (instruction-only), and all code is included in the package. No downloads or external install steps are enforced by the skill metadata. This is lower supply-chain risk than automatic fetch/install, but the README suggests cloning from a GitHub repo if users choose to do so (manual).
Credentials
The skill declares no required environment variables or credentials, which matches the described purpose. One minor inconsistency: the metadata lists 'required binaries: none', yet SKILL.md and README assume Python >= 3.6 and use execute_command to run python scripts. The skill also requires the host to provide execute_command and web_fetch tools and internet access — normal for a web-scraping search tool but worth noting.
Persistence & Privilege
The skill is not always-enabled and does not request elevated platform privileges. It writes cache files under its own script directory (.cache) and generates reports locally; it does not modify other skills or global agent config. Autonomous invocation is allowed (platform default) but not set to always:true.
Assessment
This skill appears to do what it says: run multi-source searches (using web_search or a bundled scraper), score and analyze results, and generate reports. Before installing or enabling it, consider the following:
- Review the Python scripts (they're bundled) yourself or have someone you trust scan them — the agent will run them via execute_command, so they will execute code on the host environment.
- Ensure you are comfortable granting the agent execute_command and web_fetch access and permitting network access. These are necessary for the tool to perform web searches and fetch pages; they also enable arbitrary shell execution of the bundled scripts.
- Note the skill writes local cache files (.cache) and report files in the skill directory. If you want to avoid persistent data, run it in a sandbox or inspect/clear the cache folder after use.
- The SKILL.md assumes a working python3 binary (>=3.6) but the skill metadata did not list python as a required binary; ensure python is available if you plan to use the fallback scripts.
- If you are concerned about supply-chain risk, avoid cloning the GitHub URL in the README automatically; instead copy the provided code bundle from a trusted source or review the repository before use.
If you want higher assurance, run the skill in an isolated environment (container/VM) first and verify network behavior, file writes, and that prompts/AI calls are consistent with expected outputs.Like a lobster shell, security has layers — review code before you run it.
latestvk977fh8e1sbjw5ysz53fd0tnnd82vthk
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
