Building swarm observability: traces, message grids and weather replay for agent swarms
strickvl · x · 2026-10-31
strickvl shares a toolkit for "swarm observability" to help AI engineers debug agent swarms:
- Collector: tracks everything agents do — filesystem reads/writes, inter-agent messages, and traditional traces.
- Viewer modes: "weather replay" (each agent is a dot colored by its current activity, scrubbable at different speeds), contact sheets per team, inter-agent message grids, file-edit heatmaps with clickable diffs, git-push timelines, and side-by-side run comparisons resembling classic ML experiment trackers.
- hodhod agent: a simple agent with deterministic tools to slice through the collected data and build explanatory views answering questions like what the swarm did or where it misbehaved.
The author notes it works best on smaller swarms so far, with complex interactions still an open problem; the core idea is standardized visual explanations built like HTML artifacts.
Related event: Developer Builds Mantiq for Observing 10,000-Agent Swarms(3 posts)→
More from coding & agent
- 174-line mental schema + Claude Code: an agent workflow that builds entire slide decks — tianshi_li · 2026-11-01
- Immunologist uses OpenAI Codex to simulate single-cell drug experiments — DeryaTR_ · 2026-11-01
- Stripe Link goes live in Agent Picnic to manage AI agent connections and permissions — jeff_weinstein · 2026-11-01
- BossBot books pickleball courts at midnight daily, beating the reservation rush — chrisfleck · 2026-11-01
- OpenAI's Astra 5.5 runs massive code patches just to list Devcon side events — devanshmehta · 2026-11-01
- Curated skills collection lets you run agents on your own harness and subscriptions — jasonkneen · 2026-11-01