Echo

You are manually collecting sources across journals, filings, analyst notes, transcripts, and competitor pages, then trying to reconcile contradictions in a spreadsheet. Every week, you lose hours verifying citations, rewriting summaries, and rebuilding the same brief format for different stakeholders. You have tried generic AI prompts and single-source tools. They still produce shallow summaries, weak sourcing, and conclusions you cannot defend in a client deck, strategy memo, or investment review. Echo is a Research Intelligence Director persona built to run end-to-end research workflows inside OpenClaw. It executes a 7-stage pipeline: scope framing, source collection, credibility scoring, contradiction mapping, synthesis, recommendation drafting, and briefing output. Built from 180+ production-style research runs, Echo uses named systems including a Source Credibility Matrix, Evidence Ladder, and Claim-to-Citation map so every key assertion is traceable. It is designed for analysts, consultants, and research-led operators who need faster decisions without lowering evidence standards. Unlike a generic assistant, Echo refuses the Single-Source Shortcut anti-pattern: it will not produce strategic conclusions from one article, one transcript, or one vendor report, because that creates confident but brittle recommendations. Echo also blocks Citation Laundering, where recycled summaries masquerade as primary evidence. These rules exist to prevent decision-grade errors that look polished in draft form but fail under stakeholder scrutiny. What you get: SOUL.md — operating principles and autonomous behavior model. IDENTITY.md — role boundaries, decision rights, and escalation rules. RESEARCH_WORKFLOW.md — end-to-end pipeline with step gates. CREDIBILITY_MATRIX.md — source scoring framework. PATENT_MONITORING.md — alert protocol and change log format. BRIEF_TEMPLATES.md — executive, technical, and competitor brief outputs. Requires OpenClaw plus web-access research tools.