{"data":{"event":{"id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","slug":"collaborative-memory-for-multi-agent-vlm-systems-b414786464","title":"Collaborative Memory for Multi-Agent VLM Systems","short_summary":"arXiv:2609.17921v1 Announce Type: new \nAbstract: Vision-language model (VLM) agents combine specialized perception, tools, and reasoning to address complex visual tasks. In multi-agent settings, different agents inspect different image regions, video frames, or visual representations, so collaboration extends beyond distributed reasoning to distributed perception. This makes shared visual context a central problem in VLM agent collaboration. In this paper, we frame memory hierarchy, cross-agent sharing, and consistency mechanisms around the need to reconcile interpretations and update dependen","full_description":"arXiv:2609.17921v1 Announce Type: new \nAbstract: Vision-language model (VLM) agents combine specialized perception, tools, and reasoning to address complex visual tasks. In multi-agent settings, different agents inspect different image regions, video frames, or visual representations, so collaboration extends beyond distributed reasoning to distributed perception. This makes shared visual context a central problem in VLM agent collaboration. In this paper, we frame memory hierarchy, cross-agent sharing, and consistency mechanisms around the need to reconcile interpretations and update dependent reasoning. Effective collaboration requires agents to build on contributions from other agents, recover missing visual context, and reconcile differing interpretations as new evidence emerges. Shared visual memory preserves not only images or textual summaries but also the dependencies among observations, agent interpretations, and subsequent reasoning. Together, these design considerations shape how information flows and evolves across VLM agents. The proposed framework provides a foundation for building reliable and resource-efficient agent teams.\n\nUpdate (2026-09-17): Launch HN: Skillsync (YC W26) – AI chat sessions made portable across agents. Hey HN, we're Nars & Nishant, founders of Skillsync (https://skillsync.com)\nSkillsync lets you move your AI chats across every coding agent. Most of our work exists as conversations, which are currently scattered across our agents. Though stored locally, these conversations use different formats. This is annoying because you cannot simply switch between agents without starting over. We get locked into a single provider and their agent as we invest in skills and memories over time. Skillsync acts as a universal converter. It moves the entire session, including all the messages, reasoning and tool calls so you can pick up right where you left off.\nSkillsync collects all your sessions in one place and makes them searchable. It breaks down what each session is carrying, including which loaded skills the agent is actually using, making stale context easy to spot. You can also create shared workspaces to sync sessions across your team. You can build your own closed loop systems. Everything runs locally except when you share to workspaces.\nThe core is an open-source Rust engine called txcr\n\nUpdate (2026-09-17): Launch HN: Skillsync (YC W26) – AI chat sessions made portable across agents. Hey HN, we're Nars & Nishant, founders of Skillsync (https://skillsync.com)\nSkillsync lets you move your AI chats across every coding agent. Most of our work exists as conversations, which are currently scattered across our agents. Though stored locally, these conversations use different formats. This is annoying because you cannot simply switch between agents without starting over. We get locked into a single provider and their agent as we invest in skills and memories over time. Skillsync acts as a universal converter. It moves the entire session, including all the messages, reasoning and tool calls so you can pick up right where you left off.\nSkillsync collects all your sessions in one place and makes them searchable. It breaks down what each session is carrying, including which loaded skills the agent is actually using, making stale context easy to spot. You can also create shared workspaces to sync sessions across your team. You can build your own closed loop systems. Everything runs locally except when you share to workspaces.\nThe core is an open-source Rust engine called txcr\n\nUpdate (2026-09-21): How V7 gives AI agents institutional memory. Using GPT-5.6, V7 turns scattered company files into context agents can use to complete complex, source-linked work.","ledger_type":"benefit","primary_domain_id":"fc7d617f-c9df-4b9f-b595-4408915ea380","event_status":"provisional","event_date":"2026-09-17T00:00:00.000Z","discovery_date":"2026-09-17T00:00:00.000Z","first_published_date":"2026-09-17T00:00:00.000Z","last_reviewed_date":"2026-09-21T00:00:00.000Z","geographic_scope":"International","affected_population":null,"base_impact_tier":1,"base_score":"1.00","attribution_multiplier":"0.1000","evidence_multiplier":"0.4000","realization_multiplier":"0.2000","durability_multiplier":"0.5000","current_event_score":"0.004000","confidence_level":"low","score_explanation":"Ingest update: independent source raised evidence; named Claude 3.7 Sonnet, OpenAI Codex; added a follow-up source. Score remains provisional until review. Follow-up ingest: independent source raised evidence; added a follow-up source. Follow-up ingest: independent source raised evidence; added a follow-up source.","methodology_version_id":"8745541b-0f8e-4d5a-9b47-5291d3813a8d","original_methodology_version_id":"8745541b-0f8e-4d5a-9b47-5291d3813a8d","published_at":"2026-09-17T04:00:52.788Z","created_at":"2026-09-17T04:00:52.788Z","updated_at":"2026-09-21T14:01:12.460Z","flags":[],"domain_name":"Other","domain_slug":"other","methodology_version":"0.2"},"contributions":[{"id":"d8f08c9a-c3bb-48a1-9405-d1ca93ec763c","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","model_id":"d96eca85-24db-4c1d-9a28-ca66dc5a2847","role_description":"Remapped from unspecified after ingest named OpenAI Codex.","attribution_multiplier":"0.1000","credit_share":"0.5000","contribution_score":"0.002000","attribution_rationale":"Follow-up source named OpenAI Codex.","attribution_confidence":"medium","first_used_date":null,"model_version_if_known":null,"review_status":"approved","created_at":"2026-09-17T17:01:04.189Z","model_slug":"openai-codex","model_name":"OpenAI Codex","identity_class":"named_release","is_internal":false,"family_name":"Codex","family_slug":"openai-codex","organization_name":"OpenAI","organization_slug":"openai"},{"id":"f8deb902-7bf9-4f4c-a1cf-5881d681ae37","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","model_id":"570f77f8-5233-4c57-9ce5-7753b481beee","role_description":"Remapped from unspecified after ingest named Claude 3.7 Sonnet.","attribution_multiplier":"0.1000","credit_share":"0.5000","contribution_score":"0.002000","attribution_rationale":"Follow-up source named Claude 3.7 Sonnet.","attribution_confidence":"medium","first_used_date":null,"model_version_if_known":null,"review_status":"approved","created_at":"2026-09-17T17:01:04.173Z","model_slug":"claude-3-7-sonnet","model_name":"Claude 3.7 Sonnet","identity_class":"named_release","is_internal":false,"family_name":"Claude","family_slug":"anthropic-claude","organization_name":"Anthropic","organization_slug":"anthropic"}],"sources":[{"id":"71ec8ea7-058f-414c-aa44-792586f47802","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","url":"https://arxiv.org/abs/2609.17921","canonical_url":"https://arxiv.org/abs/2609.17921","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-17T00:00:00.000Z","retrieved_at":"2026-09-17T04:00:52.801Z","title":"Collaborative Memory for Multi-Agent VLM Systems","excerpt":"arXiv:2609.17921v1 Announce Type: new \nAbstract: Vision-language model (VLM) agents combine specialized perception, tools, and reasoning to address complex visual tasks. In multi-agent settings, different agents inspect different image regions, video frames, or visual representations, so collaboration extends beyond distributed reasoning to distributed perception. This makes shared visual context a central problem in VLM agent collaboration. In this paper, we frame memory hierarchy, cross-agent ","content_hash":null,"source_reliability_class":"medium","is_primary_source":true,"is_independent":true,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-17T04:00:52.801Z"},{"id":"fac2cedc-af37-4d79-9ab7-599da2207163","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","url":"https://skillsync.com","canonical_url":"https://skillsync.com","source_type":"journalism","publisher":"Hacker News","author":null,"publication_date":"2026-09-17T00:00:00.000Z","retrieved_at":"2026-09-17T17:01:04.147Z","title":"Launch HN: Skillsync (YC W26) – AI chat sessions made portable across agents","excerpt":"Hey HN, we're Nars & Nishant, founders of Skillsync (https://skillsync.com)\nSkillsync lets you move your AI chats across every coding agent. Most of our work exists as conversations, which are currently scattered across our agents. Though stored locally, these conversations use different formats. This is annoying because you cannot simply switch between agents without starting over. We get locked into a single provider and their agent as we invest in skills and memories over time. Skillsync acts","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":true,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-17T17:01:04.147Z"},{"id":"6128038e-813c-47b3-9afb-6883953c546f","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","url":"https://news.ycombinator.com/item?id=49743049","canonical_url":"https://news.ycombinator.com/item?id=49743049","source_type":"journalism","publisher":"Hacker News","author":null,"publication_date":"2026-09-17T00:00:00.000Z","retrieved_at":"2026-09-17T18:00:52.369Z","title":"Launch HN: Skillsync (YC W26) – AI chat sessions made portable across agents","excerpt":"Hey HN, we're Nars & Nishant, founders of Skillsync (https://skillsync.com)\nSkillsync lets you move your AI chats across every coding agent. Most of our work exists as conversations, which are currently scattered across our agents. Though stored locally, these conversations use different formats. This is annoying because you cannot simply switch between agents without starting over. We get locked into a single provider and their agent as we invest in skills and memories over time. Skillsync acts","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":true,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-17T18:00:52.369Z"},{"id":"9d12aa2f-516c-40c7-97dd-da6494ec99a4","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","url":"https://openai.com/index/v7","canonical_url":"https://openai.com/index/v7","source_type":"developer_announcement","publisher":"OpenAI News","author":null,"publication_date":"2026-09-21T00:00:00.000Z","retrieved_at":"2026-09-21T14:01:12.398Z","title":"How V7 gives AI agents institutional memory","excerpt":"Using GPT-5.6, V7 turns scattered company files into context agents can use to complete complex, source-linked work.","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":true,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-21T14:01:12.398Z"}],"claims":[{"id":"33b52e31-4868-48a4-8f3a-694297870daf","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","claim_text":"Collaborative Memory for Multi-Agent VLM Systems","claim_type":"outcome","claim_status":"supported","confidence_score":null,"created_at":"2026-09-17T04:00:52.808Z","updated_at":"2026-09-17T04:00:52.808Z"}],"revisions":[{"id":"e2037319-d747-4dd7-aa52-14709fd842a9","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","model_id":null,"previous_score":"0.000000","new_score":"0.001000","previous_factors":{},"new_factors":{"tier":1,"evidence":0.1,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Auto-published from news ingest.","trigger_type":"news_ingest","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-17T04:00:52.814Z"},{"id":"a25d7136-95c6-4e45-bc7b-9d40e26467f7","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","model_id":null,"previous_score":"0.001000","new_score":"0.002000","previous_factors":{"tier":1,"evidence":0.1,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["Claude 3.7 Sonnet","OpenAI Codex"],"evidence":0.2,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; named Claude 3.7 Sonnet, OpenAI Codex; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["fac2cedc-af37-4d79-9ab7-599da2207163"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-17T17:01:04.260Z"},{"id":"2c2ebf80-e7ab-41f8-8608-708179294a37","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","model_id":null,"previous_score":"0.002000","new_score":"0.003000","previous_factors":{"tier":1,"evidence":0.2,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["Claude 3.7 Sonnet","OpenAI Codex"],"evidence":0.3,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["6128038e-813c-47b3-9afb-6883953c546f"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-17T18:00:52.421Z"},{"id":"d149d23d-8725-4f63-bf37-dbc185b55fd2","event_id":"3a81af2b-4141-4ebb-90ff-1a44175b569e","model_id":null,"previous_score":"0.003000","new_score":"0.004000","previous_factors":{"tier":1,"evidence":0.3,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["GPT-5.6"],"evidence":0.4,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["9d12aa2f-516c-40c7-97dd-da6494ec99a4"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-21T14:01:12.474Z"}],"secondary":[]},"methodology_version":"0.1"}