hermes-admin/hermes-agent
1
0
Fork 0
Code Issues Pull Requests Actions 4 Packages Projects Releases Wiki Activity

plugin-catalog: bump memex8 to v1.1.0

Browse Source 
Bumps the pinned SHA from b5e87b7b (v1.0.0) to 47fbe5b8 (v1.1.0) so
users running 'hermes plugins install memex8' get the six new
features:

  - per-session injection dedup + char-budget gate (Feature 1)
  - decision/evaluation memory-type promotion (Feature 2)
  - prompt-injection sanitization on read (Feature 3)
  - configurable system-prompt template (Feature 4)
  - export/import memories as JSON (Feature 5)
  - /memex8 slash command with 9 subcommands (Feature 6)

All new behaviors default OFF, so existing users see no behavior
change. The pinned image URL moves to match the new SHA so the
catalog preview still renders.

The 'provides_hooks' block is restored as on_session_end +
on_pre_compress (was empty in the prior catalog entry — both hooks
are still implemented and surfaced in plugin.yaml).

Plan: PLAN-memex8-v1.1-features.md (in the plugin repo).
This commit is contained in:
Marc
2026-09-27 15:55:51 -06:00
committed by Teknium
parent 97ac7cae9c
commit 2029a219c7
1 changed files with 7 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
plugin-catalog/memex8.yaml
View File

@@ -1,18 +1,20 @@
name: memex8
repo: https://github.com/Ex8-ca/memex8
sha: b5e87b7b9a537ec663dd0c5d68a5ebf819976abc
sha: 47fbe5b8813e712c8f0a52c6d4e5ab37e09498f5
subdir: plugins/memex8
version: "1.0.0"
description: Self-hosted vector memory (Qdrant + TurboQuant compression) with auto-discovered knowledge realms, semantic search across conversations and ingested documents, and a slumber pipeline for memory consolidation. Disclosure — auto-stores a truncated copy (first 300 chars) of every user and assistant turn, plus per-message previews of the whole transcript at session end and before context compression, to the memex8 server at MEMEX8_BASE_URL (self-hosted; default http://localhost:8080) using MEMEX8_API_KEY.
version: "1.1.0"
description: Self-hosted vector memory (Qdrant + TurboQuant compression) with auto-discovered knowledge realms, semantic search across conversations and ingested documents, and a slumber pipeline for memory consolidation. v1.1.0 adds optional per-session injection dedup with char-budget gate, decision/evaluation memory-type promotion (slower decay for load-bearing turns), prompt-injection sanitization on read, configurable system-prompt template, export/import memories as JSON, and a /memex8 slash command. All new behaviors default off. Disclosure — auto-stores a truncated copy (first 300 chars) of every user and assistant turn, plus per-message previews of the whole transcript at session end and before context compression, to the memex8 server at MEMEX8_BASE_URL (self-hosted; default http://localhost:8080) using MEMEX8_API_KEY.
maintainer: Ex8-ca
tier: community
category: memory
requires_hermes: ">=0.21.4"
docs_url: https://github.com/Ex8-ca/memex8/blob/main/plugins/memex8/README.md
image: https://raw.githubusercontent.com/Ex8-ca/memex8/b5e87b7b9a537ec663dd0c5d68a5ebf819976abc/plugins/memex8/docs/banner.png
image: https://raw.githubusercontent.com/Ex8-ca/memex8/47fbe5b8813e712c8f0a52c6d4e5ab37e09498f5/plugins/memex8/docs/banner.png
capabilities:
provides_tools: []
provides_hooks: []
provides_hooks:
- on_session_end
- on_pre_compress
provides_middleware: []
requires_env:
- MEMEX8_API_KEY