Competitive Intelligence

skills/competitive-research/references/first-run-baseline.md

First Dry-Run Baseline (2026-06-16)

Environment

  • Host: Chowmes (Hostinger VPS, Docker container)
  • Python: /opt/hermes/.venv/bin/python (3.13.5, pyyaml 6.0.3)
  • Parallel CLI: 0.7.1, /usr/local/bin/parallel-cli, authed via PARALLEL_API_KEY
  • Hermes CLI: /opt/hermes/.venv/bin/hermes
  • Model: DeepSeek V4 Pro (workhorse), Claude Sonnet 4.6 (synthesis)

Search Queries Run (15 total)

Competitor Moves (6 queries, 30 results)

  • Coveo AI search announcement release 2026
  • Bloomreach discovery commerce latest news
  • Constructor ecommerce AI search product update
  • Google Vertex AI Search retail ecommerce new feature
  • Elasticsearch enterprise search announcement
  • Meilisearch Typesense open source search release

Industry Signals (5 queries, 25 results)

  • search and product discovery trends ecommerce 2026
  • conversational commerce AI shopping agent
  • retail ecommerce site search best practices
  • Gartner Forrester IDC search discovery analyst
  • MCP A2A UCP agent protocol commerce search

AI Threats (4 queries, 20 results)

  • Perplexity shopping commerce search
  • ChatGPT search browse ecommerce product discovery
  • Google Universal Cart agentic commerce
  • AI agents replacing site search ecommerce 2026

Results

  • Monitor alerts: 0 (monitors not yet created)
  • Total findings: 75
  • Dry-run: completed in ~3 minutes
  • Synthesis: blocked by [ERROR] Hermes CLI not found (script used bare hermes instead of /opt/hermes/.venv/bin/hermes) — fixed in daily-research-run.py with HERMES_BIN constant

Raw findings saved to

/opt/data/knowledge/obsidian/MyOS/Projects/Competitive Intelligence/artifacts/competitive-research/raw/2026-06-16.json