A transparent sequence of governed evidence skills that transforms a plain-language research question into an audit-ready evidence synthesis output.
Each skill has a defined purpose, source inputs, standards alignment, human-review gate, quality checks, and traceable outputs.
A sequential chain of eleven governed evidence skills. Each skill produces a structured, traceable output that feeds the next. No skill is a black box — inputs, outputs, quality checks, and governance rules are defined and logged for every run.
A dedicated safety signal skill using pharmacovigilance disproportionality methods. Operates in parallel with the core chain. Output requires mandatory caveats and human review before any external communication.
These rules apply to every skill execution. They are not optional and cannot be overridden by user configuration.
This layer converts evidence synthesis outputs into reusable Medical Affairs workflow skills. Two skills are live today (MSL Briefing, Payer Evidence Summary). Three have their core logic already built but currently locked inside a single chain rather than exposed as a standalone, chain-agnostic skill — see the note on each card. The remainder are still on the roadmap. All skills follow the same governance model as the Core Evidence Skill Chain: defined inputs, outputs, quality checks, human-review gates, and audit logs.
Each skill writes its output to an immutable audit log. Inspect every decision, every source citation, and every quality flag — from the first database query to the final evidence summary.
Open research console → View audit trail