Issues

• fix(summarize): hardcoded 120s SUMMARY_TIMEOUT_MS aborts multi-chunk sessions on local LLMs with summary_timeout

  rohitg00/agentmemory#1686 · Filter by author RichiBad opened on Nov 5, 2026

• procedural tier can never produce a single procedure: frequency is derived from sessionIds, which is always empty

  rohitg00/agentmemory#1685 · Filter by author hptitan opened on Nov 5, 2026

• Is there any standalone offline installer?

  rohitg00/agentmemory#1654 · Filter by author JohnHerry opened on Nov 3, 2026

• Data not loaded after upgrade to 0.9.31 / iii 0.25.0 — state_store.db files present but API/viewer show 0

  rohitg00/agentmemory#1652 · Filter by author NikEnPrin opened on Nov 2, 2026
  1

• Contributor PR queue (weekly)

  rohitg00/agentmemory#1650 · Filter by author devin-ai-integration opened on Nov 2, 2026

• [Feature] Add Muse Code (muse CLI) support to agentmemory connect + auto-capture hooks

  rohitg00/agentmemory#1635 · Filter by author adi-venturo opened on Oct 29, 2026

• graph-index migrate always pauses with an opaque "storage failure" at the edges phase (20,091 nodes / 32,881 edges)

  rohitg00/agentmemory#1617 · Filter by author misterboo opened on Oct 26, 2026
  3

• memory_heal has no orphan-facet sweep, and facets/remove needs the full targetId+dimension+value triple

  rohitg00/agentmemory#1614 · Filter by author 7889545 opened on Oct 25, 2026

• targetLayer is accepted by /agentmemory/smart-search but has no effect

  rohitg00/agentmemory#1608 · Filter by author fgfg163 opened on Oct 24, 2026
  1

• feat(graph): support opt-in full provenance retention with bounded search indexes

  rohitg00/agentmemory#1604 · Filter by author masa6161 opened on Oct 23, 2026

• mem::compress fires one LLM call per observation with no batching/throttle - saturates self-hosted gateway, fail+retry spiral (production numbers)

  rohitg00/agentmemory#1603 · Filter by author herjarsa opened on Oct 22, 2026

• Hooks cut captured text before redaction, so a <private> block across the limit is stored

  rohitg00/agentmemory#1600 · Filter by author dmazhukov opened on Oct 20, 2026

• smart-search returns "Invocation stopped" on a large store while the worker keeps re-registering

  rohitg00/agentmemory#1593 · Filter by author chanfox7 opened on Oct 20, 2026

• Daemon picks an LLM provider for a placeholder key that doctor reports as missing

  rohitg00/agentmemory#1586 · Filter by author dmazhukov opened on Oct 19, 2026

• Engine responses over the ws default 100 MiB frame limit tear down the worker connection; state::list has no pagination to bound whole-scope reads

  rohitg00/agentmemory#1585 · Filter by author 7889545 opened on Oct 19, 2026
  7

• No store migration path: engine 0.22.1 cannot serve a 0.11.2 store; 15s readiness gate aborts; memory_export capped at 15 MiB

  rohitg00/agentmemory#1584 · Filter by author leopold-dauvergne opened on Oct 18, 2026
  1

• Worker drops its iii connection when an endpoint response exceeds 16 MiB (worker→engine frame cap); semantic/insights provenance lists are unbounded

  rohitg00/agentmemory#1579 · Filter by author Fguarien opened on Oct 17, 2026
  3

• Observation retention: an age limit that works without LLM compression

  rohitg00/agentmemory#1566 · Filter by author pablo-brown-rodriguez opened on Oct 14, 2026

• smart_search: saved memories are capped at 3 by session diversification, and small limits shrink the fusion pool

  rohitg00/agentmemory#1565 · Filter by author pablo-brown-rodriguez opened on Oct 14, 2026

• Feature Request: Update endpoint for projectScope on existing memories

  rohitg00/agentmemory#1564 · Filter by author a4spire opened on Oct 14, 2026

• feat: record which user and team wrote each observation and memory (client-side AGENTMEMORY_USER_ID / AGENTMEMORY_TEAM_ID)

  rohitg00/agentmemory#1549 · Filter by author espensgr opened on Oct 13, 2026

• Insight records reach 1.1 MB each: unbounded concept clusters hold 84.8% of a 465 MB mem:insights collection

  rohitg00/agentmemory#1548 · Filter by author 7889545 opened on Oct 13, 2026

• feat: AGENTMEMORY_HEADERS — send custom headers from clients to a remote agentmemory server (Cloudflare Access, auth proxies)

  rohitg00/agentmemory#1547 · Filter by author espensgr opened on Oct 13, 2026

• Search latency ~1.1 s steady-state (~4 s after idle) vs 0.6–17 ms in the project's own benchmark [corrected]

  rohitg00/agentmemory#1537 · Filter by author olegmaisak opened on Oct 12, 2026
  2

• Feature: jcode adapter support

  rohitg00/agentmemory#1528 · Filter by author dev-hann opened on Oct 12, 2026