Setup instructions for Sarup, a Thai-focused MCP server that compresses and retrieves conversation context. MCP is a standard that lets AI clients connect to external tools or data.

Find the Sarup repository, install it on macOS, Linux, WSL, or Windows, register the server, optionally install local models, and verify its compression tools.

It helps install and connect Sarup while keeping Claude Code usable if the server is unavailable.

sarup-setup — Skill for Claude Code and Codex by PHUICMT: Setup instructions for Sarup, a Thai-focused MCP server that…

Install

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.

agentmods

npx agentmods add skills/phuicmt/sarup/sarup-setup

Any agent

npx skills add PHUICMT/sarup --skill sarup-setup

Clone the repo

git clone --depth 1 https://github.com/PHUICMT/sarup

Made for: Claude Code, Codex.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/phuicmt/sarup/sarup-setup.svg)](https://agentmods.dev/skills/phuicmt/sarup/sarup-setup)

Your own site

<a href="https://agentmods.dev/skills/phuicmt/sarup/sarup-setup"><img src="https://agentmods.dev/badge/skills/phuicmt/sarup/sarup-setup.svg" alt="Measured on agentmods" height="20"></a>

Per session 62 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 969 The whole file, excluding the scripts and references it only reads on demand.

Security scan B 1 finding. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00062	$0.00969
Opus 5	$0.00031	$0.00485
Sonnet 5	$0.00012	$0.00194
Haiku 4.5	$0.00006	$0.00097

Measured 3d ago against content hash 9d8647d9024b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

sarup-setup scanned grade B with 1 finding 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 3d 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

`<project-dir>/.claude/settings.json`; the MCP stays user-scoped.

skills/sarup-setup/SKILL.md · 80 lines

How it starts

The opening of the file, as written. The whole thing — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Sarup setup

Set up Sarup — a Thai-first context-compression MCP server — for Claude Code. Sarup is an MCP server (tools sarup_compress / sarup_retrieve / sarup_stats), not an API proxy, so it can never break Claude Code: if it's down the tools just disappear and Claude keeps working.

Work through these steps. Stop and report if a step fails.

1. Locate the Sarup repo

Find the clone directory (it contains pyproject.toml with name = "sarup" and a scripts/ folder). In order:

If $SARUP_HOME is set, use it.
If the current working directory is the Sarup repo, use it.
Otherwise ask the user for the path. If they don't have it cloned, offer: git clone https://github.com/PHUICMT/sarup.git

2. Run setup (creates venv, installs, registers the MCP at user scope)

From the repo root:

Windows (PowerShell): ./scripts/setup.ps1 -All
Linux / WSL / macOS: ./scripts/setup.sh --all

-All / --all also pulls the optional Ollama models (nomic-embed-text, gemma3:12b) for the higher-ratio semantic / pipeline modes. Drop the flag for a minimal MCP-only install (offline extractive mode still works with no Ollama).

If the scripts can't run, do it manually with the repo's venv python:

<venv-python> -m pip install -e <repo>
claude mcp add sarup --scope user --env "SARUP_DB_PATH=<home>/.sarup-cache.db" -- <venv-python> -m sarup.server

(Windows venv python: <repo>/.venv/Scripts/python.exe; Unix: <repo>/.venv/bin/python.)

3. Verify

Run claude mcp list and confirm sarup shows Connected. Tell the user to reload Claude Code (the VS Code window or a new CLI session) so the tools load, then they can call sarup_compress(content, mode="auto") in any project.

4. (Optional) auto-compression hook

Only offer this if the user asks for automatic compression. The PostToolUse hook compresses large tool outputs (≥ SARUP_HOOK_MIN_TOKENS, default 400) and needs Claude Code ≥ 2.1.186.

⚠️ It compresses large Bash/Grep output, which can drop detail the model needs verbatim (recoverable via sarup_retrieve, but inconvenient mid-task). So prefer installing it per-project in documentation / Thai-heavy projects, and skip it in code-heavy projects. Ask the user which project before installing.

Read the full file on GitHub · 80 lines

Changes

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.

3d ago First seen · 80 lines · 62 tokens per session scan B 9d8647d9024b

Subscribe to this mod's changes

sarup-setup is a skill published in the GitHub repository PHUICMT/sarup (0 stars, last pushed 2mo ago), licensed MIT. It adds 62 tokens to every session and 969 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Building agents LLM Memory and context claude-code context-compression mcp nlp pythainlp

mem0-vercel-ai-sdk

Mem0 provider for Vercel AI SDK (@mem0/vercel-ai-provider). TRIGGER when: user mentions "vercel ai sdk", "@mem0/vercel-ai-provider", "createMem0", "retrieveMemories", "addMemories", "getMemories", "searchMemories", "mem0 vercel", "AI SDK provider", "AI SDK memory", or is using generateText/streamText with mem0. Also…

mem0ai/mem0 · 146 tokens

mem0-tour

Browses all stored memories grouped by category with full content display. Use when reviewing all project memories, exploring stored knowledge, onboarding to a project, or getting an overview of captured decisions, conventions, and learnings.

mem0ai/mem0 · 47 tokens

stats

Displays memory usage statistics for the current session and project including counts by category, age distribution, and API latency. Use when checking how many memories exist, reviewing session activity, or auditing memory distribution across categories.

mem0ai/mem0 · 43 tokens

sarup-setup