2026-10-11 18:01 UTC

Pullboard's agentic workflow with persistent institutional knowledge stored outside the codebase (items, shouts, doctrine) becomes a reference pattern for long-running agent memory and organizational learning.

state: seedheat: lowuncertainty: mediumconvergesscott: highagentic-workflow institutional-knowledge agent-memoryOlscore

What is this?

Pullboard is an open-source agentic workflow system (Show HN launch) that stores persistent institutional knowledge โ€” items, shouts, doctrine โ€” outside the codebase, aiming to give agents durable cross-session memory and organizational learning. The web results confirm the broader pattern (agent memory, institutional knowledge, cross-agent organizational memory) is actively discussed across Augment Code, Mem0, Mastra, AAIF, and academic work, but none of the returned snippets mention Pullboard specifically, so the system's exact architecture and adoption remain unconfirmed by this search.

Why it matters to Scott

Pullboard's Show HN launch โ€” an open-source agentic workflow storing 'items, shouts, doctrine' outside the codebase for durable cross-session memory and organizational learning โ€” independently arrives at the exact architecture Scott has been framing for years: stateless workers + a stateful external kernel (long-running-agents), a queryable wiki-graph as the durable kernel (wiki-is-the-kernel), institution-owned semantic memory (institutional-memory), checkpoint discipline, and handover-note hygiene. If Pullboard gains traction as a reference implementation, it becomes a dated-receipts opportunity for Scott's frameworks.
ip:framework.long-running-agentsip:framework.wiki-is-the-kernelip:concept.institutional-memoryip:concept.kernelip:framework.cognitive-gitip:concept.checkpoint-disciplinedev:project.dev-wikidev:technology.prime-agentdev:concept.resumable-agent-job-control-planedev:concept.llm-navigated-wikiradar:1f916-persistent-agent-worldradar:ablo-multi-agent-shared-stateradar:49ide-spatial-agent-workspaceradar:aafp-commons-signed-agent-notebook
queries asked of Scott's wikis
  • agent memory persistence outside codebase
  • institutional knowledge storage for coding agents
  • open-source agent harnesses and workflow systems
  • cross-session agent memory patterns
  • organizational learning in agent systems
  • durable knowledge for long-running coding agents

Measured heat

now 0 pts/hpeak 2 pts/hcomments 0/hpeers p16momentum: steady1 platformsage 45h
points/hour across evidence ยท reading as of 2026-10-12 02:59:37.977291+11:00 ยท deterministic, not a model opinion

How the heat travelled

10-09 18:46โญ origin directly observedShow HN: Pullboard โ€“ An open source agentic workflow
Olscore on hacker news
โ€”
10-09 18:46amplified on hacker news ๐Ÿ‘‘hn.story.50025023
Olscore
peak 2 ยท 0 comments ยท 98% of case engagement
10-09 19:34our radar first saw it ยท +0.8hdiscovery anchor: hn.story.50025023โ€”
pace: p33 vs 968 stories at the 24h mark (now 45h old) โ€” ahead of aafp-commons-signed-agent-notebook (2.0x), behind agentlane-git-native-coordination (0.7x)

Evidence (1) โ€” โญ canonical anchor

sourceobjectauthorscorecomments
๐ŸŸง hn โญShow HN: Pullboard โ€“ An open source agentic workflowOlscore20

Interpretation history

Decision trace