Reference Checker Skill

Exhaustively verify manuscript references before journal submission. Use when checking whether references are real, accurate, complete, and formatted consist...

This skill appears coherent and instruction-only, but consider the following before installing or running it: (1) It assumes the agent can perform web searches or access bibliographic APIs (PubMed, Crossref, publisher pages). If your agent environment blocks outbound browsing/search tools, provide the metadata or allow web access. (2) Exhaustive mode can generate many web lookups and may be slow or trigger rate limits; for very long lists consider batching or explicit sampling. (3) Paywalled content or ambiguous/older records may still require human/manual checks — the skill already marks such cases as 'Manual check'. (4) Do not feed the skill sensitive documents you don't want external lookups performed on; although the skill doesn't request credentials, it will query external sites to verify references. (5) If you plan to integrate with specific APIs (publisher APIs, Crossref), consider providing appropriate API keys or tools through your agent in a controlled way rather than embedding secrets into the skill.

Type: OpenClaw Skill Name: reference-checker Version: 1.0.0 The reference-checker skill is a prompt-based tool designed to verify academic citations for accuracy and authenticity. The instructions in SKILL.md and README.md focus entirely on bibliographic auditing, metadata verification (DOI, PubMed, Crossref), and reporting formatting errors or potential fabrications. There is no evidence of malicious intent, data exfiltration, or unauthorized command execution; the skill explicitly states it has no Python dependencies and operates purely through structured prompting.