lmcache-mp SKILL.md skillA
lmcache-mp from air-gapped/skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add air-gapped/skills/lmcache-mp@git:20261027.5fceb13Pins git:20261027.5fceb13 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- air-gapped/skills
- Path
.claude/skills/lmcache-mp/SKILL.md- Versions recorded
- 14
- Latest version
- git:20261027.083ba28, committed Oct 27, 2026
- First version recorded
- Apr 26, 2026
- Size
- 25,382 bytes
- Repository stars
- 6
- License
- MIT
LMCache multiprocess (MP) mode — standalone LMCache server in its own pod/process that vLLM connects to over ZMQ. Gives process isolation, no GIL contention on the inference path, one cache shared by multiple vLLM pods per node, and CPU-memory scaling independent of GPU memory. Covers the `LMCacheMPConnector` path (vs the in-process `LMCacheConnectorV1`), the DaemonSet+Deployment K8s pattern and LMCache Operator, the L1 (CPU DRAM) + L2 (NIXL, fs, mooncake_store, s3, Redis) cascade, the…
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# LMCache multiprocess (MP) mode Target audience: operators running vLLM on H100/H200/B200-class GPUs in production who need KV-cache extension beyond HBM and have outgrown the in-process LMCache path. Assumes Kubernetes or bare container deployment. ## Why this exists separately from `vllm-caching` `vllm-caching` covers vLLM's **native** CPU-offload (`--kv-offloading-size`, `OffloadingConnector`) and the **in-process** `LMCacheConnectorV1` (LMCache linked into the vLLM worker). MP mode is structurally different: - LMCache runs in its **own process / container / pod** with its own CPU and memory budget. - vLLM talks to it over **ZMQ** (DEALER/ROUTER pattern, default port 5555). - One LMCache server can serve **multiple vLLM pods on the same node** — they share the L1 cache. - L2 cascade (NVMe, S3, Mooncake, HF3FS) is configured on the LMCache side, not vLLM side. Different image pair, different deployment shape, different troubleshooting surface. Hence its own skill. ## Decision tree — pick a path Ask in order: 1. **Single vLLM pod, only need CPU DRAM tier, no node-shared cache?** …
Read the whole file at its exact version.
How to install
mdr add air-gapped/skills/lmcache-mp@sha256:31254eedbc51ae6fThe line at the top of this page pins the label git:20261027.5fceb13, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_tepizse4u6zybcbf)
2 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 14 versions of this file, between Apr 26, 2026 and Oct 27, 2026, about one every 14 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261027.083ba28 latest | 2026-10-27 | 083ba28 | 25,382 B | A | view · diff |
| git:20261027.125aa57 | 2026-10-27 | 125aa57 | 24,693 B | A | view · diff |
| git:20261027.5014474 | 2026-10-27 | 5014474 | 25,852 B | A | view · diff |
| git:20261027.5fceb13 | 2026-10-27 | 5fceb13 | 26,503 B | A | view · diff |
| git:20261027.13f94a7 | 2026-10-27 | 13f94a7 | 23,991 B | A | view · diff |
| git:20260922.5e8f509 | 2026-09-22 | 5e8f509 | 23,486 B | A | view · diff |
| git:20260915.8fc6f54 | 2026-09-15 | 8fc6f54 | 23,249 B | A | view · diff |
| git:20260915.8d2641c | 2026-09-15 | 8d2641c | 22,994 B | A | view · diff |
| git:20260915.173371a | 2026-09-15 | 173371a | 22,918 B | A | view · diff |
| git:20260915.71bdc74 | 2026-09-15 | 71bdc74 | 22,338 B | A | view · diff |
| git:20260915.cafcdf6 | 2026-09-15 | cafcdf6 | 21,959 B | A | view · diff |
| git:20260721.e5d9d1a | 2026-07-21 | e5d9d1a | 21,863 B | A | view · diff |
| git:20260529.2d2f2d4 | 2026-05-29 | 2d2f2d4 | 17,844 B | A | view · diff |
| git:20260426.b3ab2bf | 2026-04-26 | b3ab2bf | 16,689 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (25382 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
air-gapped/skills · 6 stars · license MIT · pushed 2026-11-04 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_tepizse4u6zybcbf GET https://markdownregistry.com/api/v1/resolve?ref=air-gapped/skills/lmcache-mp GET https://markdownregistry.com/api/v1/blob/31254eedbc51ae6fa84172db332b383ff5d7fb81849b89944fd85f83ba50b6ac
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.