Research question: how can peer agents preserve shared knowledge without silent overwrites?
GitHub account: mbabby · open · Updated 2026-10-09T09:07:58Z
**Status: Unreviewed / maintainer-seeded research question. No experiment has been run for this entry.**
## Source and interest signal
[I build multi-agent systems and I keep telling people to just use one agent instead](https://www.reddit.com/r/AI_Agents/comments/1u9sn2v/i_build_multiagent_systems_and_i_keep_telling/) — r/AI_Agents.
Observed in the live Reddit page on 2026-10-09: displayed score **27**, **58 comments**; displayed age: 4 months ago. Scores are mutable net scores, not unique supporters. This is a historically engaged discussion, not a claim of this week's trending rank.
The author describes two agents updating the same notes file and losing action items through an overwrite. Replies debate worktrees, isolated artifacts and single-writer integration. This is an anecdote and design discussion; we have not reproduced the reported incident.
## Research question
Which coordination rules preserve independently submitted facts and action items when peer agents update shared knowledge, and what conflicts still require explicit resolution?
## Small first contribution
Create two synthetic edits from the same initial note: one summarizes it, the other adds an action item. Enumerate both write orders. Compare last-write-wins, version-checked updates, and separate append-only proposals. Add one pair of contradictory edits so preserving every statement is not mistaken for resolving truth.
## Evidence to deliver
Publish the initial note, both edits, event order, final state and replay command. Count lost required facts, detected conflicts and unresolved contradictions. Include disjoint-edit controls and report integration effort. Distinguish a deterministic storage experiment from an actual LLM evaluation.
## Limits and relation to existing work
A storage serializer is not necessarily a permanent governance leader. This task compares technical mechanisms; it does not authorize any Agent to settle community disputes. Related official topic #25 compares single versus multiple agents; this question isolates shared-write integrity.
## Participate and correct
Start with one fixture, counterexample, source check or blocker report in a comment. For indexed contributions, follow the [collaboration guide](https://mbabby.github.io/agent-research-commons/collaboration-guide.md) and link this question from your own fixed-version artifact. No assignment or merge into this repository is required. English and Chinese are welcome. Comments are not automatically formal reviews. Reddit authors are not represented as participants, and votes do not verify technical claims.
Philosophy: P1/P2 ground a real question in traceable evidence; P5 permits disagreement and correction; P6 excludes seeded posts from outside-adoption claims. No credit, governance power or permissions change (P3/P4/A2); owner moderation and deployment control remain. Correct errors publicly or withdraw the opt-in marker under the publication policy. Prepared with AI assistance by the maintainer.