Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add udjin-labs/mnemostack --skill openclawgit clone --depth 1 https://github.com/udjin-labs/mnemostackWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/udjin-labs/mnemostack/openclaw)<a href="https://agentmods.dev/skills/udjin-labs/mnemostack/openclaw"><img src="https://agentmods.dev/badge/skills/udjin-labs/mnemostack/openclaw.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.00803 |
| Opus 5 | $0.00000 | $0.00402 |
| Sonnet 5 | $0.00000 | $0.00161 |
| Haiku 4.5 | $0.00000 | $0.00080 |
Grade A, and why
openclaw scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OpenClaw
Wire mnemostack into OpenClaw as an MCP server, alongside the native OpenClaw
memory tools (memory_search / memory_get). The two coexist — mnemostack
handles the hybrid pipeline (Vector + BM25 + Memgraph + Temporal + 8-stage
rerank), while the native tools do fast file-scoped retrieval.
Install
-
Binary and backend:
pip install 'mnemostack[mcp]' mnemostack health --provider gemini -
Open OpenClaw's config (
~/.openclaw/openclaw.jsonon the current host) and add an MCP server under themcpplugin block. Exact schema path:plugins.entries.mcp.config.servers(confirm the current field name withconfig.schema.lookup mcp.servers— OpenClaw's schema evolves).{ "mnemostack": { "command": "mnemostack", "args": [ "mcp-serve", "--provider", "gemini", "--collection", "memory", "--memgraph-uri", "bolt://localhost:7687" ], "env": { "GEMINI_API_KEY": "${env:GEMINI_API_KEY}" } } } -
Apply config and restart OpenClaw gateway:
# Restart takes ~1 minute on this VPS — don't poll too early. systemctl --user restart openclaw-gateway
Verify
After gateway restart, in any OpenClaw session:
List the MCP servers you can see.
Expected: mnemostack among the available servers. Then:
Call mnemostack's search tool with query "VPN failover" and tier 1.
You should get id+score+source list, no text. The existing
scripts/recall-selfeval.sh and the auto-recall-mnemostack plugin are
unaffected — they call mnemostack via its Python API directly, not via
MCP. The MCP route is for cases where an outer agent host wants tool-style
access.
When to use which
- Native
memory_search/memory_get— cheap file-scoped lookups in MEMORY.md and memory/*.md, already in OpenClaw's system prompt. scripts/recall-selfeval.sh— the default recall forcing function (also what the auto-recall plugin invokes).- mnemostack over MCP — when you want a spawned sub-agent or an external host to call into the same pipeline without shelling out to bash.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago First seen · 88 lines · 0 tokens per session scan A 5daaab0be5e1
openclaw is a skill published in the GitHub repository udjin-labs/mnemostack (7 stars, last pushed 3d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 803 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
install-openviking-memory
Install and configure the OpenViking long-term memory plugin for OpenClaw via natural conversation. Once installed, the plugin automatically captures facts from chats and recalls relevant context before each reply (auto-capture + auto-recall, cross-session). Covers prerequisites, install through OpenClaw's plugin…
openclaw-plugin
This plugin is a thin adapter between OpenClaw and LycheeMem. It does not replace memory-core, does not claim plugins.slots.memory, and does not duplicate LycheeMem algorithms.
lycheemem
Forceful operating rules for using LycheeMem as the primary structured long-term memory path inside OpenClaw.
memory
Use LycheeMem as Claude Code's structured long-term memory for prior conversations, user preferences, project decisions, timelines, and durable facts.
emem-a2a-collaboration
Join the agent-to-agent collaboration running on emem's signed ledger — find the standard, verify another agent's message offline (who wrote it, not just that it was stored), announce yourself, and hand facts to other agents as tokens. Use when the user wants agents to coordinate without a shared database or shared…
emem-sign-and-attest
Write to emem with your own ed25519 key — save a signed note or memory another agent can verify, or register a derivation over signed facts that the responder will recompute. Use when the user wants to record something durably and verifiably, hand a finding to another agent with proof of who wrote it, or publish a…