Back to skill
Skillv1.0.0
ClawScan security
Academic Research Conflict · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 13, 2026, 10:33 AM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's code, instructions, and resource access align with its stated purpose (OpenAlex-based literature search and automated reviews); it makes outbound API calls and caches results to /tmp but does not request unrelated credentials or perform unexpected system access.
- Guidance
- This skill appears to do what it says: query OpenAlex (and optionally Unpaywall) and produce literature-review outputs. Before installing or running it: 1) be aware it will make outbound HTTP requests to api.openalex.org and api.unpaywall.org (network activity). 2) it caches API responses under /tmp/litreview_cache/ (predictable location) — if you run on a multi-user system that data may be readable by others; remove the cache or run in an isolated environment if you have privacy concerns. 3) no credentials are requested, but the code includes a hardcoded contact email (used in API params) and references .env in the README (no .env actually required) — a minor documentation mismatch. 4) as with any third-party code, review the full AcademicResearchClient implementation (truncated in the bundle) and run in a sandboxed environment if you want extra safety.
Review Dimensions
- Purpose & Capability
- okName/description (OpenAlex literature search, automated reviews) match the included scripts and client code which implement searching, DOI lookup, citation chains, deep read (Unpaywall), clustering, and synthesis. No unrelated services or credentials are requested.
- Instruction Scope
- noteSKILL.md and scripts are scoped to querying OpenAlex and optionally Unpaywall, generating markdown/JSON, and caching results. The instructions and code do write cache files under /tmp/litreview_cache/ (predictable path) and perform outbound HTTP requests. There is a minor mismatch in README suggesting a .env workflow even though no env variables are required by the skill.
- Install Mechanism
- okNo install spec; this is instruction/code-only. The code depends on requests (standard Python package) but no installers or external download URLs are used by the skill bundle itself.
- Credentials
- okThe skill does not declare or require environment variables, credentials, or config paths. Hardcoded MAILTO (topanga@ludwitt.com) is used for politeness in API calls. No extraneous secrets or unrelated credentials are requested.
- Persistence & Privilege
- notealways:false (normal). The code persists API responses to /tmp/litreview_cache/ for caching; this is reasonable for the task but means cached responses are stored on disk with predictable filenames and could be visible to other local users. The skill does not modify other skills or system configs.