DNA sequencing pipelines span quality control, alignment, variant calling and annotation, and workflow managers already orchestrate mature bioinformatics tools at scale. The remaining manual component is the decision layer: setting sample‑ and platform‑specific quality thresholds, adjudicating borderline variants, diagnosing anomalies, and deciding which findings need expert review. These judgments are repetitive, operator‑dependent, inconsistent, and often undocumented.\ \ BaseCamp introduces a framework of six specialized agents that handle sample intake & QC, alignment, variant calling, annotation, cross‑stage monitoring and reporting. Crucially, the agents do not perform sequence analysis; established tools still execute alignment, calling and annotation. The agents select among tools, generate configurations, interpret outputs and decide subsequent actions, confining LLM reasoning to the reliable judgment layer while preserving the reproducibility guarantees of existing software.\ \ Agent reasoning is powered by a consortium of fine‑tuned, domain‑specific large language models coordinated by a central reasoning LLM, all running locally so sequencing data never leaves the environment, with human‑in‑the‑loop oversight at key points. Evaluation shows agent‑generated configurations match expert practice, an explicit filtering ledger makes otherwise hidden removals traceable, and cross‑stage anomaly detection surfaces issues missed by stage‑wise monitoring. BaseCamp offers a generalizable blueprint for agentic automation of scientific data pipelines.\ \ Review