# LogStrip

> Free, zero-dependency CLI that compresses noisy logs into high-signal AI context. 705+ detected ecosystems, stream-safe, deterministic output.

LogStrip is a stream-based CLI that trims noisy CI, server, scanner and container logs down to the diagnostic context an LLM needs: it scores lines, folds repeats, masks secrets and keeps JSON valid, with no model and no network. It ships as an npm CLI, a TypeScript library and a GitHub Action.

Every page below is plain Markdown. The whole documentation in one file: https://mrwogu.github.io/logstrip/llms-full.txt

## Get Started

- [Install LogStrip CLI](https://mrwogu.github.io/logstrip/getting-started/index.md): Install the LogStrip log compression CLI in one command. File, pipe, or stdin - trim noisy logs for AI agents in seconds.
- [705+ Supported Log Sources](https://mrwogu.github.io/logstrip/reference/sources/index.md): LogStrip detects 705+ log ecosystems automatically - CI, Docker, Kubernetes, Spring Boot, Trivy, Snyk, nginx, and more. Full catalogue by category.

## Evaluate

- [Why LogStrip](https://mrwogu.github.io/logstrip/why-logstrip/index.md): Logs are not prose. Why LogStrip trims logs by structure instead of dropping tokens with a language model, what it keeps on purpose, what it drops, what it rewrites, and when not to use it.
- [Benchmarks](https://mrwogu.github.io/logstrip/benchmarks/index.md): Measured token savings for every LogStrip test fixture, how the numbers are produced, what they do not tell you, and how to reproduce them with one command.
- [Compare LogStrip](https://mrwogu.github.io/logstrip/compare/index.md): How LogStrip differs from pasting raw logs, tail and grep, model-based prompt compressors such as LLMLingua, and log shippers or observability platforms. Facts per category, no invented numbers.

## Reference

- [Log Compression CLI Reference](https://mrwogu.github.io/logstrip/reference/cli/index.md): Complete CLI reference for LogStrip - flags, exit codes, I/O contract, stats, JSON output, multiline joining, severity filtering, include/exclude, timeout, progress. Extend with .logstrip.yml.
- [Core TypeScript API Reference](https://mrwogu.github.io/logstrip/reference/core/index.md): Process logs in Node.js with processLogStream and processLogFile. Stream-based, deterministic, zero external dependencies. Sanitization, scoring, dedup, multiline joining, severity filtering, format detection, and timeout API.
- [GitHub Action for Log Compression](https://mrwogu.github.io/logstrip/reference/action/index.md): Drop LogStrip into any GitHub Actions workflow with one step. Automatic Step Summary, zero config, read-only permissions. CLI is preferred for non-GitHub CI.

## Guides

- [How Log Compression Works for AI](https://mrwogu.github.io/logstrip/guides/compression-flow/index.md): Step-by-step breakdown of how LogStrip scores, sanitizes, dampens, folds, and collapses logs into high-signal AI context. Hybrid context engine, not regex-only filtering.
- [CLI and Action Configuration](https://mrwogu.github.io/logstrip/guides/configuration/index.md): Configure input, output, aggressiveness levels, stats, and JSON reporting for the LogStrip CLI and GitHub Action. Default and custom settings explained.
- [AI Agent Plugins for Log Compression](https://mrwogu.github.io/logstrip/guides/plugins/index.md): Install LogStrip plugins for Claude Code, Copilot, Cursor, Factory Droid, Codex, and OpenCode. Compress-first workflow so every agent receives smaller, higher-signal logs.
- [Log Sanitization and Security](https://mrwogu.github.io/logstrip/guides/security/index.md): How LogStrip sanitizes UUIDs, timestamps, hashes, and IPs before logs reach AI agents. Not a secret scanner - recommended security posture for CI workflows.
- [Troubleshooting Log Compression](https://mrwogu.github.io/logstrip/guides/troubleshooting/index.md): Fix empty output, exit codes, missing files, and short stack traces in the LogStrip CLI and GitHub Action. Common errors and solutions.

## Examples

- [LogStrip Cookbook](https://mrwogu.github.io/logstrip/examples/index.md): Copy-paste workflow recipes for compressing CI logs - npm/Vitest/Jest, pytest, docker compose, Kubernetes, and Jenkins. All recipes favour the CLI.
- [Compress npm test logs with LogStrip](https://mrwogu.github.io/logstrip/examples/npm-test/index.md): Trim noisy npm/yarn/pnpm/Vitest/Jest output in CI before handing it to an AI agent. Stdin pipes, matrix jobs, artifact upload, GitHub Action wrapper.
- [Compress pytest output with LogStrip](https://mrwogu.github.io/logstrip/examples/pytest/index.md): Trim pytest tracebacks and pip noise before handing logs to an AI agent. Multiline joining, GitHub Actions, GitLab CI, and stdin pipes.
- [Compress docker compose logs with LogStrip](https://mrwogu.github.io/logstrip/examples/docker-compose/index.md): Trim multi-container docker compose logs into compact, AI-ready context. Follow mode, per-service compression, CI integration.
- [Compress Kubernetes logs with LogStrip](https://mrwogu.github.io/logstrip/examples/k8s/index.md): Trim kubectl, stern, and pod logs into compact, AI-ready context. Multi-pod aggregation, crash-loop debugging, ArgoCD pipelines.
- [Compress Jenkins pipeline logs with LogStrip](https://mrwogu.github.io/logstrip/examples/jenkins/index.md): Trim noisy Jenkins console output before forwarding to an AI agent. Declarative pipelines, scripted pipelines, archiveArtifacts, post-build steps.

## Optional

- [Source code](https://github.com/mrwogu/logstrip): GitHub repository
- [npm package](https://www.npmjs.com/package/logstrip): logstrip CLI and TypeScript library
- [Changelog](https://github.com/mrwogu/logstrip/blob/main/CHANGELOG.md): release notes