ncbi-entrez-skill

Use the esearch endpoint with db=pubmed and a term parameter, such as {"endpoint":"esearch","params":{"db":"pubmed","term":"KRAS AND colorectal cancer","retmode":"json","retmax":10}}. The script returns a compact list of matching record IDs.

Use esummary or efetch with db=gene or db=protein and a specific record ID, for example {"endpoint":"esummary","params":{"db":"gene","id":"7157","retmode":"json"}}. Results are returned as compact summaries limited by max_items.

Yes, GEO is accessible through Entrez using db=gds for DataSets and series-level metadata or db=geoprofiles for profile-level records. Use GSE[ETYP] or GSM[ETYP] in search terms to filter by accession type.

Yes, the script reads NCBI_API_KEY or NCBI_EUTILS_API_KEY from environment variables and adds it to requests automatically. It also supports NCBI_TOOL and NCBI_EMAIL environment variables for NCBI usage policies.

Set save_raw=true in the input JSON and optionally provide raw_output_path. The script writes the complete response to the specified file and reports the saved path instead of returning the full payload inline.

Errors return ok=false with an error code such as invalid_input, invalid_json, network_error, or invalid_response. Common causes include missing endpoint fields, malformed JSON input, non-integer max_items values, or network failures reaching the NCBI servers.