# caveman | Julius Brussee | Knowledge Base | Instagit

🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman

GitHub Stars: 82.8k

Repository: https://github.com/JuliusBrussee/caveman

---

## Articles

### [How Caveman Achieves Token Savings on Agent Outputs and Inputs: 2 Core Mechanisms Explained](/JuliusBrussee/caveman/how-caveman-achieves-token-savings)

Discover how Caveman slashes token costs with tool schema reduction and context compaction. Learn its 2 core mechanisms for saving on agent inputs and outputs.

- Tags: deep-dive
- Published: 2026-09-06

### [Where Are the Caveman Compressor Implementations Located?](/JuliusBrussee/caveman/where-caveman-compressor-implementations-located)

Discover where Caveman compressor implementations are located in the JuliusBrussee/caveman repository. Find dedicated Go files for JSON text HTML code diff log and more within the engine compressors package.

- Tags: how-to-guide
- Published: 2026-09-06

### [How to Check Session Token Usage Statistics with Caveman: A Complete Guide](/JuliusBrussee/caveman/how-check-session-token-usage-caveman)

Easily check session token usage statistics with Caveman. Learn how the built-in caveman-stats hook displays detailed token numbers and estimated savings for your Claude Code sessions.

- Tags: how-to-guide
- Published: 2026-09-06

### [Complete List of Caveman Slash Commands in Claude Code](/JuliusBrussee/caveman/caveman-slash-commands-claude-code)

Discover all Caveman slash commands for Claude Code. Explore file operations, compression, and utility functions invoked with a leading /.

- Tags: api-reference
- Published: 2026-09-06

### [How Caveman Mode State is Managed Per Session in Claude Code](/JuliusBrussee/caveman/how-caveman-mode-state-managed-claude-code)

Discover how Caveman mode state is managed per session in Claude Code. Learn about its session-specific lifecycle and lack of cross-invocation persistence for efficient agent runs.

- Tags: internals
- Published: 2026-09-06

### [How Claude Code Hooks Integrate with Caveman: Complete Technical Guide](/JuliusBrussee/caveman/how-claude-code-hooks-integrate-caveman)

Discover how Claude Code hooks integrate with Caveman. This guide details SessionStart, UserPromptSubmit, and Statusline hooks for mode initialization, prompt tracking, and visual feedback.

- Tags: how-to-guide
- Published: 2026-09-06

### [Caveman Skill Intensity Levels: The Complete Guide (6 Modes Explained)](/JuliusBrussee/caveman/caveman-skill-intensity-levels)

Explore Caveman skill intensity levels: lite, full, ultra, wenyan-lite, wenyan-full, and wenyan-ultra. Control prose compression with this complete guide.

- Tags: deep-dive
- Published: 2026-09-06

### [Caveman Skills Explained: Structure, Configuration, and Usage in the Caveman LLM Proxy](/JuliusBrussee/caveman/what-are-caveman-skills-how-structured)

Learn how to structure, configure, and use Caveman skills. These modular plug-ins transform LLM responses and expose new commands via markdown files with YAML front-matter and plain-text rules.

- Tags: deep-dive
- Published: 2026-09-06

### [Where to Find Caveman Agent Profile Definitions: Location and Usage Guide](/JuliusBrussee/caveman/where-find-agent-profile-definitions-caveman)

Discover where to find Caveman agent profile definitions. Locate and understand these JSON files in the agents profiles directory for your AI agent configurations.

- Tags: how-to-guide
- Published: 2026-09-06

### [Caveman Wrap Command-Line Options: How to Configure Agent Sessions with Flags](/JuliusBrussee/caveman/caveman-wrap-agent-command-line-options)

Learn Caveman wrap command-line options to configure agent sessions with flags like --off --pixel --workflow and --full-auto. Control your agent's behavior efficiently.

- Tags: how-to-guide
- Published: 2026-09-06

### [How Caveman Wraps Existing Coding Agents like Claude Code and Codex: A Technical Deep-Dive](/JuliusBrussee/caveman/how-caveman-wraps-coding-agents)

Discover how Caveman wraps coding agents like Claude Code and Codex. Learn about its innovative proxy injection technique for seamless integration without code modification.

- Tags: deep-dive
- Published: 2026-09-06

### [What Are the Different Proxy Modes Available in Caveman?](/JuliusBrussee/caveman/caveman-proxy-modes)

Explore Caveman's three proxy modes: default, MCP-shrink, and legacy HTTP proxy. Learn how to optimize your caching with these flexible options.

- Tags: how-to-guide
- Published: 2026-09-06

### [How the Caveman Proxy Handles API Credentials: Architecture and Security Explained](/JuliusBrussee/caveman/how-caveman-proxy-handles-api-credentials)

Discover how the Caveman proxy secures API credentials using CAVE_API_KEY and forwarding provider-specific keys. Learn about its architecture and security measures.

- Tags: architecture
- Published: 2026-09-06

### [Which AI Model Providers Does the Caveman Proxy Support? A Complete Technical Guide](/JuliusBrussee/caveman/caveman-proxy-supported-ai-model-providers)

Discover which AI model providers Caveman proxy supports including OpenAI Azure OpenAI Anthropic Amazon Bedrock Google Vertex AI and Gemini. Access all backends with one interface.

- Tags: technical-guide
- Published: 2026-09-06

### [How to Start the Local Caveman Proxy: Installation, Startup, and Verification](/JuliusBrussee/caveman/how-to-start-local-caveman-proxy)

Learn how to start the local Caveman proxy easily. Install the CLI with npm and run `caveman start` or let it auto-launch with any agent command.

- Tags: how-to-guide
- Published: 2026-09-06

### [Caveman CCR Limitations: Security Risks, Archival Constraints, and Storage Boundaries](/JuliusBrussee/caveman/caveman-ccr-security-archival-limitations)

Discover Caveman CCR limitations: unencrypted data storage, 512 MiB limit, and security risks. Learn how its open-fail design impacts archival and exposes sensitive payloads.

- Tags: deep-dive
- Published: 2026-09-06

### [How to Retrieve Original Bytes Using Caveman CCR Handles](/JuliusBrussee/caveman/how-to-retrieve-original-bytes-caveman-ccr)

Retrieve original bytes from Caveman CCR handles using the CLI command or the TypeScript SDK. Easily recover source code with caveman mem recover.

- Tags: how-to-guide
- Published: 2026-09-06

### [Caveman CCR Storage Backends: SQLite vs In-Memory Implementation](/JuliusBrussee/caveman/caveman-ccr-storage-backends)

Explore Caveman CCR storage backends: SQLite for persistence and in-memory for WebAssembly. Seamless content compression recovery across platforms with a unified Store interface.

- Tags: deep-dive
- Published: 2026-09-06

### [What Information Is Stored by Caveman Context Recovery?](/JuliusBrussee/caveman/information-stored-by-caveman-context-recovery)

Caveman Context Recovery stores byte-exact payloads, MIME types, compressor metadata, token accounting, and auxiliary data indexed by a SHA-256 handle. Discover what CCR preserves.

- Tags: deep-dive
- Published: 2026-09-06

### [How Caveman Generates Context-Addressed Handles: SHA-256 Content Hashing Explained](/JuliusBrussee/caveman/how-caveman-handles-are-generated)

Discover how Caveman generates context-addressed handles using SHA-256 content hashing for idempotent storage and byte-exact retrieval. Learn about deterministic handle generation.

- Tags: deep-dive
- Published: 2026-09-06

### [Caveman Context Recovery (CCR): How It Preserves Exact Original Bytes in Lossy Compression](/JuliusBrussee/caveman/what-is-caveman-context-recovery)

Discover Caveman Context Recovery (CCR) an innovative content-addressable storage system that guarantees exact original bytes retrieval for lossy compressed data using SHA-256 handles.

- Tags: deep-dive
- Published: 2026-09-06

### [How Caveman Handles Inputs That Cannot Be Compressed or Become Larger After Transformation](/JuliusBrussee/caveman/caveman-handle-uncompressible-inputs)

Caveman ensures lossless compression by returning original input unchanged if transformation results in equal or larger size. Prevent token-count regression with this guarantee.

- Tags: how-to-guide
- Published: 2026-09-06

### [Caveman Compression Engine Modes: A Complete Guide to Text Compression Levels](/JuliusBrussee/caveman/caveman-compression-engine-modes)

Explore the nine Caveman compression engine modes from off to ultra including Wenyan and workflow modes. Optimize your text compression today.

- Tags: how-to-guide
- Published: 2026-09-06

### [Caveman Compression Engine Input Type Detection: JSON, Binary, and Text Classification](/JuliusBrussee/caveman/how-caveman-compression-engine-detect-input-types)

Discover how the Caveman compression engine intelligently detects input types like JSON, binary, and text using a three-step heuristic for efficient data compression. Learn more.

- Tags: internals
- Published: 2026-09-06

### [MIT/BSL-1.1 License Impact on Caveman Component Usage](/JuliusBrussee/caveman/what-is-impact-of-mit-bsl-1-1-license-on-component-usage)

Understand the MIT/BSL-1.1 license impact on JuliusBrussee/caveman component usage. Discover SDK freedom vs. core engine restrictions for your projects.

- Tags: licensing-guide
- Published: 2026-09-04

### [What Determines When a Compress Transform Fails Safe and Forwards Original Bytes](/JuliusBrussee/caveman/what-determines-when-compress-transform-fails-safe-and-forwards-original-bytes)

Learn why the compress transform fails safe and forwards original bytes due to input validation errors, sentinel conflicts, exceptions, verification mismatches, or size reduction failures.

- Tags: internals
- Published: 2026-09-04

### [How the `cav trial` Command Performs A/B Testing and Generates Comparative Reports](/JuliusBrussee/caveman/how-does-trial-command-ab-test-sessions-and-generate-comparative-reports)

Learn how the cav trial command performs A/B testing by running sequential configurations, storing telemetry, and generating side-by-side reports of latency, token usage, and model output differences.

- Tags: how-to-guide
- Published: 2026-09-04

### [How Cavecrew Presets Compress Subagent Behavior in Caveman](/JuliusBrussee/caveman/how-do-cavecrew-presets-compress-subagent-behavior)

Discover how Cavecrew presets like investigator, builder, and reviewer compress subagent behavior in Caveman, slashing token usage by 60-80% via the POST SDK v1 compress endpoint.

- Tags: internals
- Published: 2026-09-04

### [How the SSRF Allowlist Protects Against Proxy-Side Request Forgery in Caveman](/JuliusBrussee/caveman/how-does-ssrf-allowlist-protect-against-proxy-side-request-forgery)

Learn how the SSRF allowlist in Caveman safeguards against proxy-side request forgery by validating outbound URLs against unsafe addresses and an environment-driven allowlist.

- Tags: deep-dive
- Published: 2026-09-04

### [How Caveman Telemetry Reporting Works and What Data Is Never Sent](/JuliusBrussee/caveman/how-does-telemetry-reporting-work-and-what-data-is-never-sent)

Discover how Caveman telemetry reporting captures operational metadata and securely filters out sensitive data like prompts and credentials before persistence. Learn what data is never sent.

- Tags: deep-dive
- Published: 2026-09-04

### [What Compression Savings Each Detector Type Targets and Keeps in Caveman](/JuliusBrussee/caveman/what-compression-savings-does-each-detector-type-target-and-what-does-it-keep)

Discover how Caveman's detectors optimize compression for JSON, Logs, and Terminal output. Eliminate redundancy, retain semantic tokens, and boost LLM inference efficiency.

- Tags: deep-dive
- Published: 2026-09-04

### [How caveman-browse Captures Accessibility Snapshots via Chrome DevTools Protocol](/JuliusBrussee/caveman/how-does-caveman-browse-capture-accessibility-snapshots-via-chrome-devtools-protocol)

Discover how caveman-browse captures accessibility snapshots using the Chrome DevTools Protocol and the chromedp Go library to analyze your web page's accessibility tree.

- Tags: how-to-guide
- Published: 2026-09-04

### [How CacheGuard and Prefix Replacement Optimizers Work at the Provider Level in Caveman](/JuliusBrussee/caveman/how-do-optimizers-like-cacheguard-and-prefix-replacement-work-at-provider-level)

Discover how CacheGuard and prefix replacement optimizers function at the provider level in Caveman. Learn about efficient cache consistency and fast prefix substitution.

- Tags: deep-dive
- Published: 2026-09-04

### [How the Caveman Proxy Preserves Credentials While Routing Through Different Provider Adapters](/JuliusBrussee/caveman/how-does-proxy-preserve-credentials-while-routing-through-different-provider-adapters)

Learn how the Caveman proxy safeguards credentials by using a Proxy-Authorization header, stripping them before forwarding to ensure secure LLM provider routing.

- Tags: how-to-guide
- Published: 2026-09-04

### [Caveman Source Map Explained: How Engine, Proxy, MCP, and Browser Modules Interact](/JuliusBrussee/caveman/what-is-source-map-and-how-do-engine-proxy-mcp-and-browser-modules-interact)

Understand the Caveman source map. Learn how Engine, Proxy, MCP, and Browser modules interact to map output tokens to original source locations for deterministic execution and replay.

- Tags: internals
- Published: 2026-09-04

### [How the Caveman Safety Classifier Determines Which Payloads Skip Transformation](/JuliusBrussee/caveman/how-does-safety-classifier-determine-which-payloads-should-skip-transformation)

Discover how the Caveman safety classifier skips payload transformation. Learn about record mode, missing compressors, CCR requirements, and compression failures.

- Tags: how-to-guide
- Published: 2026-09-04

### [How Caveman Learn Analyzes Agent History and Ranks Token Sinks for Optimization](/JuliusBrussee/caveman/how-does-caveman-learn-analyze-agent-history-and-rank-token-sinks-for-optimization)

Discover how Caveman Learn analyzes agent history, quantifies token consumption, and ranks token sinks to optimize LLM spending and identify cost drivers.

- Tags: how-to-guide
- Published: 2026-09-04

### [How the caveman-shrink Tool Compresses Tool Catalogs with Byte-Exact Recovery](/JuliusBrussee/caveman/how-does-caveman-shrink-tool-compress-tool-catalogs-with-byte-exact-recovery)

Learn how caveman-shrink compresses tool catalogs with byte-exact recovery. This tool uses lossy schema transformation and durable storage for perfect original data restoration. Optimize your tool definitions today.

- Tags: how-to-guide
- Published: 2026-09-04

### [How Streaming Request Compression Differs from Standard Non-Streaming Transforms in Caveman](/JuliusBrussee/caveman/how-does-streaming-request-compression-differ-from-standard-non-streaming-transforms)

Discover how Caveman's streaming request compression differs from standard non-streaming transforms. Learn about MCP tool requirements and synchronous API calls for efficient data processing.

- Tags: deep-dive
- Published: 2026-09-04

### [What Happens When CCR Storage Is Unavailable or Full During Compression in Caveman](/JuliusBrussee/caveman/what-happens-when-ccr-storage-is-unavailable-or-full-during-compression)

Learn what happens when CCR storage is unavailable or full during compression in Caveman. Discover how Caveman's Shrink function ensures operational continuity by falling back to pass-through mode.

- Tags: how-to-guide
- Published: 2026-09-04

### [How Agent Wrapping Works for Claude Code, Codex, and Other LLM Agents](/JuliusBrussee/caveman/how-does-agent-wrapping-work-for-claude-code-codex-and-other-supported-agents)

Discover how Caveman's agent wrapping creates a local proxy for LLM traffic, enabling compression, metering, and tracing for Claude Code, Codex, and more without code changes.

- Tags: how-to-guide
- Published: 2026-09-04

### [How the MCP Server Exposes `caveman_retrieve` for Decompressing Payloads Inside an Agent](/JuliusBrussee/caveman/how-does-mcp-server-expose-caveman-retrieve-for-decompressing-payloads-inside-agent)

Discover how the MCP server exposes caveman_retrieve to decompress agent payloads locally via the CCR store. Get original bytes without upstream model involvement.

- Tags: how-to-guide
- Published: 2026-09-04

### [How TOON Re-encoding Compresses Text Visually for Smaller Model Input](/JuliusBrussee/caveman/how-does-toon-re-encoding-compress-text-visually-for-smaller-model-input)

Discover TOON re-encoding to compress JSON text visually, slashing token counts by 30-50% for smaller model input without losing data fidelity. Learn how it works.

- Tags: deep-dive
- Published: 2026-09-04

### [How Pixel Mode Converts Skill Text to PNG Pages and When It Fires](/JuliusBrussee/caveman/how-does-pixel-mode-convert-skill-text-to-png-pages-and-when-does-it-fire)

Discover how Caveman CLI's pixel mode converts skill text to PNG images and when it activates. Learn about this lossy compression feature for efficient skill management.

- Tags: how-to-guide
- Published: 2026-09-04

### [How contextwindow.Pack() Uses BM25 Relevance Scoring to Fit Context Into a Token Budget](/JuliusBrussee/caveman/how-does-contextwindow-pack-use-bm25-relevance-scoring-to-fit-context-into-token-budget)

Learn how contextwindow Pack uses BM25 relevance scoring to prioritize and select valuable context fragments, fitting them efficiently within your token budget.

- Tags: how-to-guide
- Published: 2026-09-04

### [Internal Request Flow Through caveman-proxy: From Agent to Provider Explained](/JuliusBrussee/caveman/what-is-internal-request-flow-through-caveman-proxy-from-agent-to-provider)

Understand the internal request flow through caveman-proxy. See how agent requests are routed from entry to provider, including authentication, routing, and adapters.

- Tags: internals
- Published: 2026-09-04

### [How the `detect()` Function Classifies Payloads and Routes Them to the Correct Compressor in Caveman](/JuliusBrussee/caveman/how-does-detect-function-classify-payloads-by-type-and-route-them-to-correct-compressor)

Discover how Caveman's detect() function classifies payloads by type and routes them to the right compressor. Learn about JSON, source code, and binary data handling.

- Tags: internals
- Published: 2026-09-04

### [How CCR SQLite Store Enables Byte-Exact Payload Recovery in Caveman](/JuliusBrussee/caveman/how-does-ccr-sqlite-store-enable-byte-exact-payload-recovery)

Learn how CCR SQLite store enables byte-exact payload recovery. It stores raw BLOBs immutably, generates deterministic handles, and retrieves data via exact-byte SELECT queries for guaranteed recovery.

- Tags: internals
- Published: 2026-09-04

### [Caveman Engine Stable Operations: The Five Core API Methods Explained](/JuliusBrussee/caveman/stable-operations-caveman-engine)

Explore the five core API methods Compress Retrieve Detect Stats and Simulate in the JuliusBrussee/caveman engine Learn about stable operations for robust integration

- Tags: api-reference
- Published: 2026-08-22

### [How to Run Tests for the Caveman Engine: A Complete Guide](/JuliusBrussee/caveman/run-tests-caveman-engine)

Easily run tests for the Caveman engine with simple commands. This guide details how to execute the full test suite covering compressors, storage, and detection logic.

- Tags: how-to-guide
- Published: 2026-08-22

### [How to Build the Caveman Compression Engine CLI: Complete Setup Guide](/JuliusBrussee/caveman/build-caveman-compression-engine-cli)

Build the Caveman compression engine CLI easily by cloning the JuliusBrussee caveman repository and running go build. Get your native binary for Compress, Detect, Retrieve, and Stats operations.

- Tags: how-to-guide
- Published: 2026-08-22

### [Natural Language Activation and Deactivation in Caveman: How the Parser Works](/JuliusBrussee/caveman/caveman-natural-language-activation-deactivation)

Discover how Caveman's regex-driven parser interprets natural language activation and deactivation commands, returning structured verdicts to control operational modes.

- Tags: internals
- Published: 2026-08-22

### [How readFlag Prevents Local Attacker Exploitation in Caveman](/JuliusBrussee/caveman/read-flag-prevent-local-attacker-exploitation)

Discover how readFlag in Caveman stops local attackers by blocking symlinks, enforcing size limits, and validating content against a whitelist. Learn more about this security feature.

- Tags: deep-dive
- Published: 2026-08-22

### [How to Configure the Default Caveman Mode: Complete Configuration Guide](/JuliusBrussee/caveman/configure-default-caveman-mode)

Learn how to configure the default Caveman mode with this complete guide. Understand the four-step resolution chain prioritizing environment variables, local configs, user settings, and built-in defaults.

- Tags: how-to-guide
- Published: 2026-08-22

### [How the UserPromptSubmit Hook Manages Caveman Activation in Claude Code](/JuliusBrussee/caveman/userpromptsubmit-hook-manage-caveman-activation)

Discover how the UserPromptSubmit hook in JuliusBrussee/caveman manages Caveman activation by parsing commands and persisting states to .caveman-active.

- Tags: internals
- Published: 2026-08-22

### [How the SessionStart Hook Initializes Sessions in Caveman: Execution Flow and Architecture](/JuliusBrussee/caveman/sessionstart-hook-interaction-caveman)

Understand how the SessionStart hook initializes sessions in Caveman. Learn its execution flow and architecture for preloading configuration and data into session context.

- Tags: architecture
- Published: 2026-08-22

### [What Is the .caveman-active Flag File in Caveman?](/JuliusBrussee/caveman/purpose-of-caveman-active-flag-file)

Discover the purpose of the .caveman-active flag file in Caveman. Learn how it controls the tool-chain's active state and operating mode for efficient workflows.

- Tags: explanation
- Published: 2026-08-22

### [How Agents Are Auto-Discovered by the Claude Code Plugin in Caveman](/JuliusBrussee/caveman/how-agents-are-auto-discovered-claude-code-plugin)

Discover how the Claude Code plugin auto-discovers agents by scanning agent markdown files in the agents directory. Learn how Caveman simplifies agent registration for seamless integration.

- Tags: how-to-guide
- Published: 2026-08-22

### [How the cavecrew-reviewer Agent Indicates Severity of Findings in Caveman](/JuliusBrussee/caveman/cavecrew-reviewer-severity-indication)

Understand how the cavecrew-reviewer agent signals finding severity with its four-tier emoji system. Easily scan bug, risk, nit, and question indicators.

- Tags: how-to-guide
- Published: 2026-08-22

### [Understanding the Limitations of the cavecrew-builder Agent in JuliusBrussee/caveman](/JuliusBrussee/caveman/limitations-of-cavecrew-builder-agent)

Discover the limitations of the cavecrew-builder agent. Learn why it avoids new files, destructive changes, and complex edits to stay within its token budget.

- Tags: deep-dive
- Published: 2026-08-22

### [Cavecrew-Investigator Agent Output Format: Structure and Syntax Rules](/JuliusBrussee/caveman/cavecrew-investigator-output-format)

Discover the cavecrew-investigator agent output format. Learn its strict token-efficient syntax rules for code locations file paths symbols and notes to understand structured data.

- Tags: api-reference
- Published: 2026-08-22

### [What Are the Cavecrew Subagents? A Technical Guide to Caveman’s AI Agent Architecture](/JuliusBrussee/caveman/purpose-of-cavecrew-subagents)

Discover Cavecrew subagents: reviewer, builder, and investigator. Learn how these AI agents enhance Caveman's code analysis, editing, and navigation for developers.

- Tags: technical-guide
- Published: 2026-08-22

### [How to Use the caveman-compress Skill to Compress a File](/JuliusBrussee/caveman/how-to-use-caveman-compress-skill)

Learn to compress files with caveman-compress. Shrink markdown or text files into terse caveman style for LLM prompts, preserving all technical content. Run caveman-compress <filepath>.

- Tags: how-to-guide
- Published: 2026-08-22

### [How the Caveman-Review Skill Formats Comments: Syntax Guide and Examples](/JuliusBrussee/caveman/caveman-review-skill-comment-format)

Learn how the caveman-review skill formats comments with its strict syntax L<line>: <problem>. <fix>. Discover how to use severity emojis and avoid filler text.

- Tags: how-to-guide
- Published: 2026-08-22

### [How to Use the caveman-commit Skill for Conventional Commit Generation](/JuliusBrussee/caveman/how-to-use-caveman-commit-skill)

Learn how to use caveman-commit to effortlessly generate Conventional Commit messages from staged diffs or natural language prompts. Streamline your Git workflow today.

- Tags: how-to-guide
- Published: 2026-08-22

### [Caveman Skill Intensity Levels: The Complete Guide to Text Compression Modes in JuliusBrussee/caveman](/JuliusBrussee/caveman/caveman-skill-intensity-levels-available)

Explore Caveman skill intensity levels: lite, full, ultra, wenyan-lite, wenyan-full, and wenyan-ultra. Control output compression from professional conciseness to extreme Classical Chinese brevity.

- Tags: deep-dive
- Published: 2026-08-22

### [How to Adjust the Intensity Level of the Caveman Skill: A Complete Guide](/JuliusBrussee/caveman/adjust-caveman-skill-intensity-level)

Easily adjust the Caveman skill intensity using slash commands like lite full or ultra. Learn how to update the flag file for session start customization.

- Tags: how-to-guide
- Published: 2026-08-22

### [Understanding the Core Caveman Skill's Terse Prose Mode: Rules and Implementation](/JuliusBrussee/caveman/caveman-skill-terse-prose-rules)

Discover the rules for Caveman skill's terse prose mode. Learn how to compress LLM responses by removing fluff while retaining all technical substance. Activate with a simple sentinel.

- Tags: how-to-guide
- Published: 2026-08-22

### [How the Caveman Skill Compresses Agent Output: A Deep Dive into the Compression Pipeline](/JuliusBrussee/caveman/how-caveman-skill-compresses-agent-output)

Discover how the Caveman skill compresses agent output. Learn about its pipeline: prose detection, front-matter stripping, prompt engineering, and validation for efficient markdown.

- Tags: deep-dive
- Published: 2026-08-22

### [Caveman Skill vs Engine/Proxy: Understanding the Architecture](/JuliusBrussee/caveman/difference-between-caveman-skill-and-engine-proxy)

Understand the Caveman skill vs Engine Proxy difference. Caveman instructs LLM terse style. Engine Proxy compresses byte payloads. Learn how they work.

- Tags: architecture
- Published: 2026-08-22

### [How to Install the Caveman Skill: Local and Global Setup Guide](/JuliusBrussee/caveman/how-to-install-caveman-skill)

Easily install the Caveman skill locally or globally. Follow our quick guide to set up Caveman in your current repository or for machine-wide access.

- Tags: how-to-guide
- Published: 2026-08-22

### [How to Install caveman on POSIX and Windows Systems](/JuliusBrussee/caveman/caveman-installation-posix-windows)

Easily install caveman on POSIX and Windows systems using simple curl or PowerShell one-liners. Get your AI agent configured with plugin support quickly and efficiently.

- Tags: getting-started
- Published: 2026-07-13

### [How caveman-shrink MCP Server Performs Compression Tasks](/JuliusBrussee/caveman/how-mcp-servers-caveman-shrink-perform-compression)

Discover how the caveman-shrink MCP server optimizes text by normalizing whitespace removing stop words shortening URLs masking secrets and preserving code blocks Get efficient data compression.

- Tags: internals
- Published: 2026-07-13

### [How caveman-commit Manages Breaking Changes and Migration Notes in Commit Messages](/JuliusBrussee/caveman/caveman-commit-handling-breaking-changes-migration-notes)

Learn how caveman-commit ensures breaking changes and migration notes are documented in commit messages. It enforces a mandatory commit body for impact details and migration steps.

- Tags: how-to-guide
- Published: 2026-07-13

### [Caveman Hook Scripts Explained: Activate, Statusline, and Mode-Tracker Triggers](/JuliusBrussee/caveman/caveman-hook-scripts-and-triggers)

Learn about Caveman hook scripts activate statusline and mode-tracker. Discover how these scripts trigger automatically to enhance your coding workflow with Claude.

- Tags: deep-dive
- Published: 2026-07-13

### [How the OpenClaw Library Integrates with the Caveman System](/JuliusBrussee/caveman/openclaw-library-integration-caveman-system)

Discover how the OpenClaw library integrates seamlessly with the Caveman system. Learn about skill definitions and bootstrap blocks managed by openclaw.js for continuous operation.

- Tags: integration-guide
- Published: 2026-07-13

### [How caveman-compress Processes Files Containing Both Code and Prose](/JuliusBrussee/caveman/how-caveman-compress-handles-mixed-content-files)

Discover how caveman-compress efficiently processes mixed content files. It separates code and prose, minifies only executable snippets, and preserves documentation readability.

- Tags: how-to-guide
- Published: 2026-07-13

### [Understanding the Severity Prefixes (🔴/🟡/🔵/❓) in caveman-review Comments](/JuliusBrussee/caveman/meaning-of-severity-prefixes-caveman-review)

Discover the meaning behind caveman-review tool's severity prefixes 🔴🟡🔵❓ Understand code review findings instantly with these emoji indicators for bugs, risks, nits, and questions.

- Tags: api-reference
- Published: 2026-07-13

### [Caveman Privacy Implications: What Data Never Leaves Your Machine](/JuliusBrussee/caveman/what-are-the-privacy-implications-and-what-data-does-caveman-never-transmit)

Discover Caveman's privacy implications. Learn what data never leaves your machine with this privacy-first tool that transmits zero user data post-installation.

- Tags: deep-dive
- Published: 2026-07-12

### [How Caveman Integrates with Claude Code, Cursor, and Cline AI Coding Agents](/JuliusBrussee/caveman/how-does-caveman-interact-with-different-ai-coding-agents-claude-code-cursor-cline)

Discover how Caveman integrates with Claude Code, Cursor, and Cline AI coding agents. Learn about its dual-architecture system for efficient, filler-free communication.

- Tags: how-to-guide
- Published: 2026-07-12

### [cavecrew-reviewer Output Format and Sorting Rules in Caveman](/JuliusBrussee/caveman/what-is-the-cavecrew-reviewer-output-format-and-how-does-it-sort-findings)

Discover the cavecrew-reviewer output format and sorting rules for clear, concise findings. Learn how findings are sorted alphabetically by file and numerically by line number.

- Tags: api-reference
- Published: 2026-07-12

### [The Difference Between Cavecrew and Vanilla Explore Agents: Context Usage Explained](/JuliusBrussee/caveman/what-is-the-difference-between-cavecrew-and-vanilla-explore-agents-in-context-usage)

Understand the context usage difference between Cavecrew and vanilla Explore agents. Cavecrew agents use 65% less context, enabling complex workflows and preventing context window exhaustion.

- Tags: deep-dive
- Published: 2026-07-12

### [How cavecrew-builder Signals “too-big” or “ambiguous” Tasks in Caveman](/JuliusBrussee/caveman/how-does-cavecrew-builder-signal-when-a-task-is-too-big-or-ambiguous)

Learn how cavecrew-builder signals too-big or ambiguous tasks with terminal refusal lines, enabling parent agents to split work or seek clarification efficiently.

- Tags: deep-dive
- Published: 2026-07-12

### [What Is the Caveman Auto-Clarity Feature and When Does It Disable Itself for Safety?](/JuliusBrussee/caveman/what-is-the-auto-clarity-feature-and-when-does-caveman-disable-itself-for-safety)

Discover the Caveman auto-clarity feature, a CLI tool safety mechanism that compresses LLM outputs and its safety disabling conditions. Learn when it turns off automatically.

- Tags: deep-dive
- Published: 2026-07-12

### [How Wenyan Levels Achieve Additional Compression Using Classical Chinese in Caveman](/JuliusBrussee/caveman/how-do-the-wenyan-levels-achieve-additional-compression-using-classical-chinese)

Discover how Wenyan levels in JuliusBrussee/caveman achieve 80-90% compression using Classical Chinese grammar. Learn how this reduces tokens for efficient data handling.

- Tags: how-to-guide
- Published: 2026-07-12

### [What Happens When caveman-compress Hits a Retry Failure After 2 Attempts](/JuliusBrussee/caveman/what-happens-when-caveman-compress-encounters-a-retry-failure-after-2-attempts)

Discover what caveman-compress does after two retry failures. Learn how it restores original files and signals compression failure.

- Tags: internals
- Published: 2026-07-12

### [How Caveman Handles Multi-Language Conversations: Style Compression Without Translation](/JuliusBrussee/caveman/how-does-caveman-handle-multi-language-conversations-while-compressing-style-but-not-translating)

Learn how Caveman achieves multi-language conversation compression without translation. It preserves original language, URLs, and code blocks, maintaining linguistic identity in your markdown. Read more!

- Tags: deep-dive
- Published: 2026-07-12

### [Security Considerations When Using Caveman with Sensitive Code: Local-First Architecture Explained](/JuliusBrussee/caveman/what-security-considerations-exist-when-using-caveman-with-sensitive-code)

Learn security considerations for using Caveman with sensitive code. Explore local-first architecture, input validation, and prompt injection risks for JuliusBrussee/caveman.

- Tags: security-guide
- Published: 2026-07-12

### [How cavecrew-investigator Formats Output for Grep Parsing](/JuliusBrussee/caveman/how-does-cavecrew-investigator-format-its-output-to-enable-grep-parsing)

Learn how cavecrew-investigator formats output for grep parsing using a simple line-based structure. Process code symbol searches efficiently with Unix tools without complex JSON.

- Tags: how-to-guide
- Published: 2026-07-12

### [What Hooks Does Caveman Install for Automatic Activation on Claude Code?](/JuliusBrussee/caveman/what-hooks-does-caveman-install-for-automatic-activation-on-claude-code)

Discover the seven hook scripts installed by Caveman for automatic activation on Claude Code. Learn how Caveman registers for SessionStart and UserPromptSubmit events.

- Tags: internals
- Published: 2026-07-12

### [How Token Savings Is Calculated in the /caveman-stats Command: A Complete Technical Breakdown](/JuliusBrussee/caveman/how-is-token-savings-calculated-in-the-caveman-stats-command)

Learn how token savings are calculated in the caveman stats command. This technical breakdown details output token counting, compression ratios, and USD conversion.

- Tags: deep-dive
- Published: 2026-07-12

### [What Triggers Cause Caveman to Revert to Normal Mode Automatically](/JuliusBrussee/caveman/what-triggers-cause-caveman-to-revert-to-normal-mode-automatically)

Discover the automatic triggers that revert Caveman to normal mode. Learn about deactivation commands, specific phrasing, one-shot commands, and new sessions in this technical guide.

- Tags: how-to-guide
- Published: 2026-07-12

### [How Does the caveman-shrink MCP Middleware Compress Tool Descriptions?](/JuliusBrussee/caveman/how-does-the-caveman-shrink-mcp-middleware-compress-tool-descriptions)

Discover how caveman-shrink MCP middleware compresses tool descriptions. Learn its token-saving strategy by removing filler words while preserving essential code and paths.

- Tags: deep-dive
- Published: 2026-07-12

### [What File Types Can caveman-compress Process and Which Are Explicitly Excluded](/JuliusBrussee/caveman/what-file-types-can-caveman-compress-process-and-which-are-explicitly-excluded)

Discover which file types caveman-compress supports, including .md and .txt, while learning which are excluded like code files to safeguard your projects.

- Tags: api-reference
- Published: 2026-07-12

### [Caveman Intensity Levels Explained: Lite, Full, and Ultra Modes](/JuliusBrussee/caveman/what-are-the-differences-between-the-lite-full-and-ultra-intensity-levels-in-caveman)

Understand caveman intensity levels lite full and ultra. Learn how each mode transforms text compression for maximum terseness and efficiency in your responses.

- Tags: deep-dive
- Published: 2026-07-12

### [Internal Architecture of Cavecrew Subagents: 3-Agent Pipeline for Context Efficiency](/JuliusBrussee/caveman/what-is-the-internal-architecture-of-the-cavecrew-subagents-and-how-do-they-reduce-context-usage)

Explore the internal architecture of Cavecrew subagents. Learn how the 3-agent pipeline optimizes context efficiency, reducing token usage by 60% with specialized roles and compressed output.

- Tags: internals
- Published: 2026-07-12

### [Does Caveman Compress Input Tokens or Only Output Tokens?](/JuliusBrussee/caveman/does-caveman-compress-input-tokens-or-only-output-tokens)

Discover if Caveman compresses input or output tokens. Learn how this tool optimizes Claude's context window by shrinking input files for efficient processing.

- Tags: faq
- Published: 2026-07-11

### [Exact Token Savings for Memory File Compression in Caveman](/JuliusBrussee/caveman/what-are-the-exact-token-savings-for-memory-file-compression)

Discover exact token savings for memory file compression with Caveman. Reduce input token counts by up to 59.6% on average to optimize LLM prompts.

- Tags: performance
- Published: 2026-07-11

### [Node.js Requirements for Caveman Installation: Complete Version Guide](/JuliusBrussee/caveman/what-are-the-node-js-requirements-for-caveman-installation)

Install Caveman easily by meeting its Node.js requirements. Discover the specific Node.js version needed for a smooth installation and avoid common setup issues.

- Tags: getting-started
- Published: 2026-07-11

### [Caveman Installation Commands for macOS, Linux, WSL, and Windows](/JuliusBrussee/caveman/what-are-the-installation-commands-for-caveman-on-different-os)

Easily install Caveman on macOS, Linux, WSL, and Windows with simple one-line commands. Automate AI coding agent setup and plugin installation for a seamless development workflow.

- Tags: how-to-guide
- Published: 2026-07-11

### [When Does Caveman Automatically Drop Compression via Auto-Clarity?](/JuliusBrussee/caveman/when-does-caveman-automatically-drop-compression-via-auto-clarity)

Discover when Caveman automatically drops compression via auto-clarity. Learn how it ensures clear communication by suspending compression during security warnings, irreversible ops, and ambiguity.

- Tags: deep-dive
- Published: 2026-07-11

### [How Caveman Handles Code Blocks During Compression: Protected Patterns in caveman-shrink](/JuliusBrussee/caveman/how-does-caveman-handle-code-blocks-during-compression)

Discover how Caveman protects code blocks and inline code during compression using protected patterns. Learn how to keep your code snippets safe with caveman-shrink.

- Tags: how-to-guide
- Published: 2026-07-11

### [What Content Does Caveman Preserve During Compression? A Technical Breakdown](/JuliusBrussee/caveman/what-content-does-caveman-preserve-during-compression)

Discover what content Caveman preserves during compression. Learn how it keeps code blocks, URLs, and technical terms while compressing prose.

- Tags: deep-dive
- Published: 2026-07-11

### [What Types of Content Does Caveman Remove During Compression?](/JuliusBrussee/caveman/what-types-of-content-does-caveman-remove-during-compression)

Caveman compress strips natural language prose, preserving code, URLs, file paths, and more. Learn what content Caveman removes during compression.

- Tags: explanation
- Published: 2026-07-11

### [What Are Cavecrew Subagents and How Do They Help?](/JuliusBrussee/caveman/what-are-cavecrew-subagents-and-how-do-they-help)

Explore Cavecrew subagents, specialized LLM delegates for efficient code analysis. They reduce token usage by 60% via token-efficient pipelines for read-only code location, editing, and diff review.

- Tags: how-to-guide
- Published: 2026-07-11

### [What Is the caveman-shrink MCP Server and How Does It Work?](/JuliusBrussee/caveman/what-is-the-caveman-shrink-mcp-server-for)

Discover the caveman-shrink MCP server, a stdio proxy that compresses prose fields in MCP responses to cut token use while keeping code and URLs intact. Learn how it works!

- Tags: deep-dive
- Published: 2026-07-11

### [What Happens to Original Files When Using /caveman-compress: Complete Backup Workflow Explained](/JuliusBrussee/caveman/what-happens-to-original-files-when-using-caveman-compress)

Discover what happens to original files when using caveman-compress. Learn how caveman protects your data with verified backups before compressing, ensuring a safe workflow.

- Tags: how-to-guide
- Published: 2026-07-11

### [How Does `/caveman-compress` Rewrite Memory Files: A Complete Technical Breakdown](/JuliusBrussee/caveman/how-does-caveman-compress-rewrite-memory-files)

Discover how caveman-compress rewrites memory files. Learn about secure backups, YAML preservation, Claude compression, validation, and atomic writes for technical content.

- Tags: deep-dive
- Published: 2026-07-11

### [How to Use the /caveman-stats Command to See Token Savings in Caveman](/JuliusBrussee/caveman/how-to-use-the-caveman-stats-command-to-see-token-savings)

Discover token savings with the /caveman-stats command in Claude Code. Track output tokens, cache reads, and estimate savings to optimize your usage. Learn more!

- Tags: how-to-guide
- Published: 2026-07-11

### [Caveman-Review PR Comment Format: Structure and Implementation](/JuliusBrussee/caveman/what-is-the-format-for-caveman-review-pr-comments)

Understand the /caveman-review PR comment format: L<line>: <severity> <problem>. <fix>. or LGTM. Learn about its structure and implementation in the JuliusBrussee/caveman repository.

- Tags: how-to-guide
- Published: 2026-07-11

### [How the `/caveman-commit` Command Generates Commit Messages: An LLM-Driven Approach to Conventional Commits](/JuliusBrussee/caveman/how-does-the-caveman-commit-command-generate-commit-messages)

Discover how caveman-commit uses an LLM to generate concise Conventional Commits with strict formatting, imperative mood, and short subject lines.

- Tags: how-to-guide
- Published: 2026-07-11

### [How to Stop Caveman Mode and Return to Normal Mode: Complete Guide](/JuliusBrussee/caveman/how-to-stop-caveman-mode-or-return-to-normal-mode)

Easily stop Caveman mode and return to normal. Learn the simple commands to deactivate Caveman mode and restore verbose output. Get the complete guide now.

- Tags: how-to-guide
- Published: 2026-07-11

### [How to Use the `/caveman` Command with Different Levels in JuliusBrussee/caveman](/JuliusBrussee/caveman/how-to-use-the-caveman-command-with-different-levels)

Learn how to use the /caveman command with various levels lite full ultra wenyan and wenyan-ultra to customize your Caveman experience in JuliusBrussee/caveman. Enhance your workflow today.

- Tags: how-to-guide
- Published: 2026-07-11

### [Understanding the Six Intensity Levels for Caveman Compression](/JuliusBrussee/caveman/what-are-the-different-intensity-levels-for-caveman-compression)

Explore the six intensity levels of Caveman compression from lite to wenyan-ultra. Control LLM output condensation and choose the best setting for your needs.

- Tags: deep-dive
- Published: 2026-07-11

### [How Caveman Is Installed for Different AI Agents: Complete Setup Guide](/JuliusBrussee/caveman/how-is-caveman-installed-for-different-ai-agents)

Install Caveman for AI agents with our complete setup guide. This unified installer auto-detects your AI coding agents and runs the correct installation routine for seamless integration.

- Tags: how-to-guide
- Published: 2026-07-11

### [How Caveman's Prompt Engineering Layer Works: A Deep Dive into the Three-Component Architecture](/JuliusBrussee/caveman/how-does-cavemans-prompt-engineering-layer-work)

Discover how Caveman's prompt engineering layer works with its three-component architecture. Learn how system prompts, mode tracking, and configuration inject dynamic instructions into Claude.

- Tags: deep-dive
- Published: 2026-07-11

### [Which AI Coding Agents Are Supported by Caveman? A Complete Compatibility Guide](/JuliusBrussee/caveman/which-ai-coding-agents-are-supported-by-caveman)

Discover which AI coding agents Caveman supports. Explore compatibility with Claude Code, Gemini CLI, Cursor, Copilot, and many more in this complete guide.

- Tags: compatibility-guide
- Published: 2026-07-11

### [What Is the Average Token Saving with Caveman? The 65% Benchmark Explained](/JuliusBrussee/caveman/what-is-the-average-token-saving-with-caveman)

Discover how Caveman achieves an average 65% token saving. Learn about its benchmark performance and understand the impressive compression rates for LLM outputs.

- Tags: performance
- Published: 2026-07-11

### [How Caveman Reduces AI Agent Output Tokens: Architecture and Implementation](/JuliusBrussee/caveman/how-does-caveman-reduce-ai-agent-output-tokens)

Learn how Caveman reduces AI agent output tokens by 65% using prompt templates. Preserve code and URLs while eliminating filler for efficient AI interactions.

- Tags: architecture
- Published: 2026-07-11

### [Security Checks in Caveman's readFlag Function: Filesystem Hardening Explained](/JuliusBrussee/caveman/what-are-the-security-checks-in-cavemans-readflag-function)

Explore security checks in Caveman's readFlag function. Discover how symlink detection, O_NOFOLLOW flags, and validation prevent local filesystem attacks protecting your data.

- Tags: deep-dive
- Published: 2026-07-09

### [Caveman Configuration Resolution Order: How Default Modes Are Determined](/JuliusBrussee/caveman/what-is-the-caveman-configuration-resolution-order)

Understand Caveman configuration resolution order. Discover how environment variables, repo-local, and user-level configs determine default modes, with a built-in fallback.

- Tags: how-to-guide
- Published: 2026-07-09

### [How to Set a Default Caveman Compression Level: Complete Configuration Guide](/JuliusBrussee/caveman/how-can-i-set-a-default-caveman-compression-level)

Learn how to set the default Caveman compression level. Configure automatic compression on startup using the CAVEMAN_DEFAULT_MODE environment variable or a configuration file setting.

- Tags: how-to-guide
- Published: 2026-07-09

### [How Long Does Caveman Level Selection Persist?](/JuliusBrussee/caveman/how-long-does-caveman-level-selection-persist)

Discover how long Caveman level selection persists. It remains active for your current session until you change intensity or end it.

- Tags: faq
- Published: 2026-07-09

### [How Wenyan Compression Works in Caveman: Rule-Based Classical Chinese Text Transformation](/JuliusBrussee/caveman/how-does-wenyan-compression-work-in-caveman)

Discover Wenyan compression in Caveman. Learn how rule-based classical Chinese text transformation reduces token count by 80-90% while preserving meaning without external translation models.

- Tags: how-to-guide
- Published: 2026-07-09

### [How to Switch Caveman Compression Levels: Ultra, Full, and Off Mode Commands](/JuliusBrussee/caveman/how-do-i-switch-caveman-compression-levels)

Easily switch Caveman compression levels with simple slash commands like /caveman ultra, /caveman full, or /caveman off. Control compression aggressiveness effortlessly.

- Tags: how-to-guide
- Published: 2026-07-09

### [Where to Find Caveman's Benchmark Evaluation Results](/JuliusBrussee/caveman/where-can-i-find-cavemans-benchmark-evaluation-results)

Find Caveman's benchmark evaluation results in the results.json file. Access detailed timing, accuracy, and resource usage metrics for all model configurations.

- Tags: how-to-guide
- Published: 2026-07-09

### [Caveman Input vs Output Token Savings: How /caveman-compress and Regular Mode Reduce Costs](/JuliusBrussee/caveman/input-vs-output-token-savings-caveman-compress-regular-mode)

Discover input token savings with /caveman-compress and output savings with regular mode. Learn how JuliusBrussee caveman reduces Claude API costs by compressing input and enforcing concise output.

- Tags: performance
- Published: 2026-07-08

### [How to Configure Caveman’s Statusline to Display Lifetime Token Savings](/JuliusBrussee/caveman/how-to-configure-statusline-display-lifetime-token-savings)

Learn how to configure Caveman statusline to display your lifetime token savings. Enhance Claude Code's statusline by enabling the savings badge with a simple setup.

- Tags: how-to-guide
- Published: 2026-07-08

### [How Caveman Handles Multi-Step Sequences When Fragment Order Is Critical](/JuliusBrussee/caveman/caveman-handling-multi-step-sequences-critical-fragment-order)

Discover how Caveman manages multi-step sequences with critical fragment order. Learn how it ensures accuracy by switching to explicit step-by-step output when needed.

- Tags: how-to-guide
- Published: 2026-07-08

### [How to Integrate Caveman with OpenClaw for Self-Hosted Agent Deployment](/JuliusBrussee/caveman/how-to-integrate-caveman-with-openclaw-self-hosted-agent)

Learn to integrate Caveman with OpenClaw for self-hosted agent deployment. Easily provision a Caveman skill into OpenClaw for an always on agent.

- Tags: how-to-guide
- Published: 2026-07-08

### [How to Measure Real Session Token Savings vs the Advertised 65% Reduction in Caveman](/JuliusBrussee/caveman/measure-real-caveman-token-savings-vs-advertised-honest-numbers)

Discover how to accurately measure Caveman session token savings. Compare raw API token counts to advertised reductions, accounting for input overhead, to find honest net savings. Learn more.

- Tags: how-to-guide
- Published: 2026-07-08

### [Running Multiple Caveman Intensity Levels in One Session: How Persistence Works](/JuliusBrussee/caveman/run-multiple-caveman-intensity-levels-session-persistence)

Learn how to run multiple Caveman intensity levels in one session and understand persistent settings. Maximize your Claude Code experience today.

- Tags: how-to-guide
- Published: 2026-07-08

### [How Caveman Handles Auto-Clarity When Compression Causes Ambiguity](/JuliusBrussee/caveman/how-caveman-handles-auto-clarity-with-compression-ambiguity)

Discover how Caveman's auto-clarity ensures clarity by disabling aggressive compression when output risks ambiguity, keeping important messages secure.

- Tags: deep-dive
- Published: 2026-07-08

### [Caveman vs Caveman‑Code: Understanding the Difference Between Mouth‑Only and Full Agent Modes](/JuliusBrussee/caveman/difference-between-caveman-code-full-agent-and-caveman-mouth-only)

Understand the difference between caveman-code full agent mode and caveman mouth-only mode. Learn how each approach compresses AI agent outputs and interactions for efficient use.

- Tags: deep-dive
- Published: 2026-07-08

### [How to Use `/caveman-commit` to Generate Conventional Commit Messages Under 50 Characters](/JuliusBrussee/caveman/how-to-use-caveman-commit-for-short-conventional-commits)

Learn how to use caveman-commit to generate conventional commit messages. This tool analyzes staged diffs for concise, effective commit messages under 50 characters.

- Tags: how-to-guide
- Published: 2026-07-08

### [Does Caveman Preserve Code Blocks, Error Strings, and CLI Commands Exactly?](/JuliusBrussee/caveman/does-caveman-preserve-code-blocks-error-strings-cli-commands)

Caveman ensures precise preservation of code blocks, error strings, and CLI commands. Experience byte-for-byte identity for your original Markdown content.

- Tags: deep-dive
- Published: 2026-07-08

### [What Is caveman-shrink and How Does It Compress MCP Server Tool Descriptions?](/JuliusBrussee/caveman/what-is-caveman-shrink-and-how-it-compresses-mcp-server-tool-descriptions)

Discover caveman-shrink, the MCP middleware proxy that compresses tool descriptions in JSON-RPC to minimize token usage while safeguarding code and URLs.

- Tags: deep-dive
- Published: 2026-07-08

### [How to Uninstall Caveman from an AI Agent: Complete Removal Guide](/JuliusBrussee/caveman/how-to-uninstall-caveman-from-ai-agent)

Uninstall Caveman from your AI agent easily. Follow our complete removal guide using npx or node to reverse all installer changes on Claude Code, Gemini CLI, and more.

- Tags: how-to-guide
- Published: 2026-07-08

### [Caveman Classical Chinese Wenyan Mode: Token Savings and Implementation](/JuliusBrussee/caveman/why-caveman-uses-classical-chinese-wenyan-mode-token-savings)

Discover Caveman's wenyan mode which rewrites replies in Classical Chinese reducing output tokens by 70-85% while preserving code blocks and technical identifiers.

- Tags: deep-dive
- Published: 2026-07-08

### [How Caveman's Hook Architecture Enables Auto-Activation on Claude Code](/JuliusBrussee/caveman/how-caveman-hook-architecture-enables-auto-activation-claude-code)

Discover how Caveman's hook architecture auto-activates on Claude Code. Learn about session start hooks, intensity detection, and rule context injection for seamless integration.

- Tags: architecture
- Published: 2026-07-08

### [What Is Cavecrew and How Do Caveman’s Sub‑Agents (Investigator, Builder, Reviewer) Work?](/JuliusBrussee/caveman/what-is-cavecrew-and-how-caveman-subagents-work)

Understand Cavecrew and its Caveman sub-agents Investigator, Builder, and Reviewer. Learn how these agents optimize token usage for efficient task delegation.

- Tags: deep-dive
- Published: 2026-07-08

### [How Caveman Handles Security Warnings and Irreversible Action Confirmations](/JuliusBrussee/caveman/how-caveman-handles-security-warnings-and-irreversible-actions)

Learn how Caveman secures critical actions by disabling token compression during security warnings and irreversible action confirmations, ensuring safety and clarity.

- Tags: how-to-guide
- Published: 2026-07-08

### [/caveman‑compress vs Regular Caveman Mode: Token Savings Explained](/JuliusBrussee/caveman/difference-between-caveman-compress-and-regular-caveman-mode)

Understand token savings: Regular Caveman mode cuts output tokens by 65% with terse speak, while /caveman-compress reduces input tokens by 46% by shrinking memory files.

- Tags: deep-dive
- Published: 2026-07-08

### [Can Caveman Compress Text in Non-English Languages Without Translation?](/JuliusBrussee/caveman/can-caveman-compress-non-english-text-without-changing-language)

Discover if Caveman can compress non-English text without translation. Learn how Caveman's language-agnostic pipeline preserves original text while reducing bloat.

- Tags: deep-dive
- Published: 2026-07-08

### [How the /caveman-stats Command Calculates Lifetime Token Savings](/JuliusBrussee/caveman/how-caveman-stats-calculates-lifetime-token-savings)

Discover how the /caveman-stats command calculates your lifetime token savings by analyzing session snapshots, mode attribution, and compression ratios from your history log.

- Tags: how-to-guide
- Published: 2026-07-08

### [Why Caveman Skips Invented Abbreviations Like cfg/impl/req but Keeps Standard Acronyms Like DB/API/HTTP](/JuliusBrussee/caveman/why-caveman-skips-invented-abbreviations-but-keeps-standard-ones)

Discover why Caveman skips invented abbreviations like cfg/impl/req but keeps standard ones like DB/API/HTTP. Learn to improve code readability and token efficiency.

- Tags: deep-dive
- Published: 2026-07-08

### [How to Install Caveman for Claude Code, Cursor, and Other AI Agents](/JuliusBrussee/caveman/how-to-install-caveman-for-claude-code-cursor-ai-agents)

Easily install Caveman for Claude Code, Cursor, and AI agents with a universal script that auto-detects and configures your tools for seamless integration.

- Tags: how-to-guide
- Published: 2026-07-08

### [Caveman Mode Intensity Levels: Lite, Full, Ultra, and Wenyan Variants Explained](/JuliusBrussee/caveman/caveman-mode-intensity-levels-lite-full-ultra)

Explore Caveman mode intensity levels: Lite, Full, Ultra, and Wenyan variants. Discover how each level progressively compresses assistant output for concise communication.

- Tags: deep-dive
- Published: 2026-07-08

### [How Caveman's Compression Algorithm Achieves 65% Token Reduction While Preserving Technical Accuracy](/JuliusBrussee/caveman/how-caveman-compression-preserves-accuracy-and-reduces-tokens)

Discover how Caveman's compression algorithm cuts 65% of LLM tokens by compressing only text, preserving 100% technical accuracy. Learn its unique approach to efficient AI.

- Tags: deep-dive
- Published: 2026-07-08

### [How to Enable MCP Shrink Middleware in Caveman: Complete Setup Guide](/JuliusBrussee/caveman/how-to-enable-mcp-shrink-middleware-in-caveman)

Enable MCP shrink middleware in Caveman effortlessly. Use the installer flag to compress text fields, reducing token usage without code changes. Learn how now.

- Tags: how-to-guide
- Published: 2026-07-07

### [How MCP Shrink Middleware Works with Caveman: A Complete Technical Guide](/JuliusBrussee/caveman/how-does-mcp-shrink-middleware-work-with-caveman)

Discover how MCP shrink middleware seamlessly compresses prose fields with Caveman, preserving essential data like URLs and code tokens. Get the complete technical guide.

- Tags: deep-dive
- Published: 2026-07-07

### [What File Types Does Caveman Compression Target? Markdown-First Architecture Explained](/JuliusBrussee/caveman/what-file-types-does-caveman-compression-target)

Caveman compression targets only Markdown files with the .md extension. Learn how it rewrites them in-place, preserving code blocks and URLs.

- Tags: architecture
- Published: 2026-07-07

### [How to Verify Caveman Installation: 5 Steps to Confirm Your Setup](/JuliusBrussee/caveman/how-to-verify-caveman-installation)

Easily verify your Caveman installation in 5 steps. Confirm agent detection, check for the .caveman-active flag, and ensure your setup is working correctly.

- Tags: how-to-guide
- Published: 2026-07-07

### [Agent-Specific Installation for Caveman: Targeted Setup Guide for AI Agents](/JuliusBrussee/caveman/agent-specific-installation-for-caveman)

Install Caveman for specific AI agents like Claude Code or Gemini CLI. Use the --only flag or agent native commands for targeted setup on JuliusBrussee/caveman.

- Tags: getting-started
- Published: 2026-07-07

### [Caveman Installation Script for Windows PowerShell: Complete Setup Guide](/JuliusBrussee/caveman/caveman-installation-script-for-windows-powershell)

Install Caveman for Windows PowerShell with our complete setup guide. This script validates Node.js and ensures a smooth, one-line installation for AI coding agents.

- Tags: how-to-guide
- Published: 2026-07-07

### [Caveman Installation Script for macOS and Linux: Complete Setup Guide](/JuliusBrussee/caveman/caveman-installation-script-for-macos-linux)

Install Caveman on macOS Linux with our guide. Use the one-line remote command or clone the repo for automatic AI agent detection and hook configuration.

- Tags: how-to-guide
- Published: 2026-07-07

### [Supported AI Agents for the Caveman Installer: Complete Integration Guide](/JuliusBrussee/caveman/supported-ai-agents-for-caveman-installer)

Discover supported AI agents for the Caveman installer. Integrate over 30 AI coding agents like Copilot, Gemini, and Claude Code automatically with this comprehensive guide.

- Tags: complete-integration-guide
- Published: 2026-07-07

### [Understanding Wenyan Intensity Levels in Caveman](/JuliusBrussee/caveman/what-are-wenyan-intensity-levels-in-caveman)

Discover wenyan intensity levels in Caveman: wenyan-lite, wenyan-full, and wenyan-ultra. Compress technical text into Classical Chinese with up to 90% reduction while keeping precise meaning.

- Tags: deep-dive
- Published: 2026-07-07

### [How to Switch Caveman Intensity Levels: A Complete Guide to Text Compression Modes](/JuliusBrussee/caveman/how-to-switch-caveman-intensity-levels)

Easily switch Caveman intensity levels lite full or ultra using the /caveman command. Persist your preference with a hidden file for consistent text compression.

- Tags: how-to-guide
- Published: 2026-07-07

### [Caveman AI Communication Intensity Levels: Complete Guide to Claude Compression Modes](/JuliusBrussee/caveman/caveman-ai-communication-intensity-levels-explained)

Explore Caveman AI communication intensity levels for Claude. Understand lite, wenyan-ultra, and other compression modes to control AI verbosity and enhance efficiency. Optimize your AI interactions now.

- Tags: deep-dive
- Published: 2026-07-07

### [How to Track Token Savings with Caveman: CLI and API Guide](/JuliusBrussee/caveman/how-to-track-token-savings-with-caveman)

Track token savings with Caveman by analyzing Claude session logs. Understand cost avoidance and monetary savings using this CLI and API guide.

- Tags: how-to-guide
- Published: 2026-07-07

### [Caveman Skill for Compressed Code Review: Ultra-Terse PR Feedback](/JuliusBrussee/caveman/caveman-skill-for-compressed-code-review)

Master ultra-terse code reviews with the caveman skill. Convert verbose feedback to single-line comments, slashing token usage by 65% for efficient PR discussions.

- Tags: how-to-guide
- Published: 2026-07-07

### [How to Load Caveman Skills into AI Agents: Complete Setup Guide](/JuliusBrussee/caveman/how-to-load-caveman-skills-into-ai-agents)

Learn to load Caveman skills into AI agents. Our guide explains session-start hooks and `bin/install.js` for seamless integration with Claude Code, Opencode, and more.

- Tags: how-to-guide
- Published: 2026-07-07

### [Caveman Skill System Components: A Technical Guide to Modular AI Compression](/JuliusBrussee/caveman/what-are-the-caveman-skill-system-components)

Explore the Caveman skill system components, including self-contained metadata files and runtime hooks for token efficiency, automated commits, and sub-agent delegation. Learn more at JuliusBrussee/caveman.

- Tags: how-to-guide
- Published: 2026-07-07

### [How Caveman Reduces AI Coding Agent Output Tokens: A Technical Deep Dive](/JuliusBrussee/caveman/how-does-caveman-reduce-ai-coding-agent-output)

Discover how Caveman reduces AI coding agent output by compressing prose up to 65% while preserving essential code and instructions. Learn the technical details.

- Tags: deep-dive
- Published: 2026-07-07

### [How to Install the Caveman AI Coding Assistant: Complete Setup Guide](/JuliusBrussee/caveman/how-to-install-caveman-ai-coding-assistant)

Install the Caveman AI coding assistant quickly with a simple command. This guide provides a complete setup for macOS, Linux, and Windows, ensuring seamless integration for your AI agents.

- Tags: how-to-guide
- Published: 2026-07-07

### [Caveman's Skill File Injection System for AI Agents: A Three-Layer Architecture](/JuliusBrussee/caveman/caveman-skill-file-injection-system-architecture-ai-agents)

Explore Caveman's three-layer skill file injection system for AI agents. Learn how it manipulates project context and JSON payloads to enhance agent behavior efficiently.

- Tags: architecture
- Published: 2026-07-07

### [How Caveman Differentiates Standard Acronyms from Invented Abbreviations in Ultra Mode](/JuliusBrussee/caveman/caveman-ultra-mode-standard-acronyms-vs-invented-abbreviations)

Discover how Caveman's ultra mode differentiates standard acronyms from invented abbreviations using mode-specific rules in skill definition files for cleaner code.

- Tags: deep-dive
- Published: 2026-07-07

### [How Benchmarks Are Measured in the `benchmarks/` and `evals/` Directories](/JuliusBrussee/caveman/benchmark-measurement-benchmarks-evals-directories)

Discover how benchmarks are measured in the `benchmarks/` and `evals/` directories of the Caveman repository to quantify token savings with robust statistical methods for accurate skill comparison.

- Tags: performance
- Published: 2026-07-07

### [How the `/caveman` Command Persists Mode Across Session Turns](/JuliusBrussee/caveman/caveman-command-persistence-session-turns)

Discover how the /caveman command persists its mode across Claude session turns by atomically writing to a hidden flag file and restoring it via the mode-tracker hook. Learn more now.

- Tags: internals
- Published: 2026-07-07

### [How to Uninstall Caveman from a Specific Agent While Keeping It for Others](/JuliusBrussee/caveman/uninstall-caveman-specific-agent-keep-others)

Uninstall Caveman from a specific agent easily. Learn how to remove traces and reinstall only the agents you need with the --only flag.

- Tags: how-to-guide
- Published: 2026-07-07

### [Caveman's Privacy Architecture: Does It Send Telemetry or Phone Home?](/JuliusBrussee/caveman/caveman-privacy-architecture-telemetry-phone-home)

Discover Caveman privacy architecture and confirm it collects zero telemetry and has no phone-home features. All data processing is 100% local.

- Tags: architecture
- Published: 2026-07-07

### [How Caveman Auto-Disables for Security Warnings and Irreversible Actions](/JuliusBrussee/caveman/caveman-auto-disable-security-warnings-irreversible-actions)

Discover how Caveman's Auto-Clarity feature auto-disables for security warnings and irreversible actions by embedding rules in the skill description for Claude. Learn more.

- Tags: how-to-guide
- Published: 2026-07-07

### [Output Token Savings vs Input Token Savings in Caveman: Complete Technical Guide](/JuliusBrussee/caveman/caveman-output-token-savings-vs-input-token-savings)

Discover output token savings and input token savings in Caveman. Learn how terse speak compresses AI responses and how compress commands reduce memory file sizes. Get technical insights.

- Tags: deep-dive
- Published: 2026-07-07

### [How Caveman Installation Hooks Manage File Ownership and CI Synchronization](/JuliusBrussee/caveman/caveman-installation-hooks-file-ownership-ci-sync)

Learn how Caveman installation hooks ensure file-ownership safety and synchronize CI workflows. Discover atomic writes, UID verification, and automatic source file mirroring for seamless management.

- Tags: how-to-guide
- Published: 2026-07-07

### [What Files Are Modified by the Caveman Installer in Agent Configuration Directories?](/JuliusBrussee/caveman/caveman-installer-modified-files-agent-config-directories)

Discover which files the Caveman installer modifies in agent configuration directories. Learn about updated JSON, JS, MD, and shell script files for Claude Code, OpenCode, and Hermes.

- Tags: how-to-guide
- Published: 2026-07-07

### [How the Claude Code Hook Enables Automatic Caveman Activation from the First Message](/JuliusBrussee/caveman/claude-code-hook-automatic-caveman-activation-first-message)

Discover how the Claude Code hook automatically activates Caveman from the first message. Learn how it injects rulesets for immediate mode activation and enhanced interaction.

- Tags: deep-dive
- Published: 2026-07-07

### [How Caveman Subagents (Investigator, Builder, Reviewer) Differ from Standard Agents](/JuliusBrussee/caveman/caveman-subagents-vs-standard-agents-investigator-builder-reviewer)

Discover how Caveman subagents investigator, builder, and reviewer differ from standard agents. Learn about their specialized roles, limited toolsets, and rigid output formats.

- Tags: deep-dive
- Published: 2026-07-07

### [caveman-shrink MCP Middleware: How It Compresses Tool Descriptions for Token Efficiency](/JuliusBrussee/caveman/caveman-shrink-mcp-middleware-compress-tool-descriptions)

Discover how caveman-shrink MCP middleware compresses tool descriptions to improve token efficiency. Learn about its sentinel-based protection system for smarter API calls.

- Tags: how-to-guide
- Published: 2026-07-07

### [How caveman-compress Preserves Code Blocks and URLs When Compressing Markdown](/JuliusBrussee/caveman/caveman-compress-preserve-code-blocks-urls-text)

Discover how caveman-compress protects code blocks and URLs during Markdown compression. Learn about its unique preservation techniques.

- Tags: how-to-guide
- Published: 2026-07-07

### [How the `/caveman-stats` Command Calculates Token Savings and Costs](/JuliusBrussee/caveman/caveman-stats-calculate-token-savings-costs)

Learn how the /caveman-stats command calculates token savings and costs by analyzing logs, applying compression benchmarks, and using pricing tables for accurate USD estimates.

- Tags: how-to-guide
- Published: 2026-07-07

### [How the Caveman-Commit Command Generates Short Conventional Commit Messages](/JuliusBrussee/caveman/caveman-commit-generate-short-conventional-commit-messages)

Discover how the caveman-commit command generates short Conventional Commit messages by routing diffs through a language model with strict formatting rules. Learn more.

- Tags: how-to-guide
- Published: 2026-07-07

### [Caveman Lite, Full, Ultra, and Wenyan Intensity Levels Explained](/JuliusBrussee/caveman/caveman-intensity-level-differences-lite-full-ultra-wenyan)

Understand Caveman intensity levels: lite, full, ultra, and wenyan. Discover how each level minimizes tokens while preserving accuracy for efficient code documentation.

- Tags: deep-dive
- Published: 2026-07-07

### [How Caveman Compresses Output Tokens While Maintaining Technical Accuracy](/JuliusBrussee/caveman/how-does-caveman-compress-output-tokens-while-maintaining-technical-accuracy)

Discover how Caveman compresses output tokens for Claude. Learn its unique pipeline for preserving technical accuracy in code, URLs, and more.

- Tags: how-to-guide
- Published: 2026-07-07

### [How Does Caveman 2 Measure and Verify Token Savings?](/JuliusBrussee/caveman/how-does-caveman-2-measure-and-verify-token-savings)

Discover how Caveman 2 accurately measures and verifies token savings. Learn how it parses session transcripts and applies benchmarked compression ratios for efficient AI.

- Tags: how-to-guide
- Published: 2026-07-06

### [Why Caveman Avoids Inventing New Abbreviations Like cfg, impl, and req](/JuliusBrussee/caveman/why-caveman-does-not-invent-new-abbreviations)

Caveman rejects invented abbreviations like cfg impl and req. Learn why these shortenings offer no token savings but hinder reader comprehension and slow development.

- Tags: best-practices
- Published: 2026-07-06

### [How to Troubleshoot Caveman Installation Issues: Complete Guide for AI Coding Agents](/JuliusBrussee/caveman/how-to-troubleshoot-caveman-installation-issues)

Troubleshoot Caveman installation issues. Verify agent detection, check PATH, and use --force to fix Claude Code or other AI agent recognition problems with this complete guide.

- Tags: how-to-guide
- Published: 2026-07-06

### [What Is caveman-code and How It Differs from the caveman Project](/JuliusBrussee/caveman/what-is-caveman-code-and-how-it-differs)

Discover caveman-code, a full-stack terminal coding agent that compresses all artifacts. Learn how it differs from the caveman project, a concise LLM speech skill.

- Tags: comparison
- Published: 2026-07-06

### [Privacy Implications of Using Caveman: A Complete Local-First Security Analysis](/JuliusBrussee/caveman/what-are-the-privacy-implications-of-using-caveman)

Explore the privacy implications of using Caveman. This local-first app offers complete security for air-gapped environments with no user data shared.

- Tags: security-analysis
- Published: 2026-07-06

### [How Caveman Handles Security Warnings and Irreversible Actions: The Auto-Clarity Rule Explained](/JuliusBrussee/caveman/how-does-caveman-handle-security-warnings-and-irreversible-actions)

Discover how Caveman's Auto-Clarity Rule safeguards against security risks and irreversible actions by temporarily disabling token compression for clear, uncompromised responses.

- Tags: how-to-guide
- Published: 2026-07-06

### [Boundaries and Exceptions for Caveman Mode: A Complete Technical Guide](/JuliusBrussee/caveman/what-are-the-boundaries-and-exceptions-for-caveman-mode)

Understand caveman mode boundaries and exceptions. Learn how this session-wide style compression skill works and its safety exceptions.

- Tags: deep-dive
- Published: 2026-07-06

### [How to Use Wenyan Mode (Classical Chinese Compression) in Caveman](/JuliusBrussee/caveman/how-to-use-wenyan-mode-classical-chinese-compression)

Learn to use Wenyan mode in Caveman to compress replies into classical Chinese. Reduce token usage by up to 90% with wenyan-full, wenyan-lite, or wenyan-ultra commands.

- Tags: how-to-guide
- Published: 2026-07-06

### [AI Coding Agents Compatible with Caveman: Complete Setup Guide](/JuliusBrussee/caveman/what-ai-coding-agents-are-compatible-with-caveman)

Discover AI coding agents compatible with Caveman. Our guide details setup for Claude Code, Gemini CLI, GitHub Copilot, and 30+ others. Integrate seamlessly today.

- Tags: how-to-guide
- Published: 2026-07-06

### [How Cavecrew Subagents (Investigator, Builder, Reviewer) Work in the Caveman AI Coding Agent](/JuliusBrussee/caveman/how-do-cavecrew-subagents-work)

Explore how Cavecrew subagents (Investigator, Builder, Reviewer) enhance the Caveman AI coding agent for efficient code search, edits, and reviews with minimal token use.

- Tags: how-to-guide
- Published: 2026-07-06

### [How caveman-compress Reduces Memory File Size for Future Sessions](/JuliusBrussee/caveman/how-does-caveman-compress-reduce-memory-file-size)

Discover caveman-compress how it slashes Claude session token use by 46% by rewriting markdown memory files into compact, reusable formats. Preserve code and URLs.

- Tags: how-to-guide
- Published: 2026-07-06

### [How to Track Token Savings with the `/caveman-stats` Command](/JuliusBrussee/caveman/how-to-track-token-savings-with-caveman-stats-command)

Learn how to track token savings with the /caveman-stats command. This tool analyzes Claude Code transcripts to estimate savings using compression ratios and flexible reporting flags. Optimize your Claude usage today.

- Tags: how-to-guide
- Published: 2026-07-06

### [caveman-shrink MCP Middleware: A Complete Guide to Compressing MCP Server Responses](/JuliusBrussee/caveman/what-is-caveman-shrink-mcp-middleware-and-how-do-i-use-it)

Learn how caveman-shrink MCP middleware compresses Model Context Protocol server responses. Reduce token usage by compressing prose while preserving code and URLs. A complete guide.

- Tags: how-to-guide
- Published: 2026-07-06

### [How Caveman Handles Different Languages: Portuguese, Spanish, and Beyond](/JuliusBrussee/caveman/how-does-caveman-handle-different-languages)

Discover how Caveman supports multiple languages like Portuguese and Spanish. Learn how it preserves your dominant language using style compression for a seamless user experience.

- Tags: deep-dive
- Published: 2026-07-06

### [How caveman-commit Generates Terse Conventional Commit Messages](/JuliusBrussee/caveman/how-does-caveman-commit-generate-terse-conventional-commit-messages)

Discover how caveman-commit generates terse conventional commit messages by enforcing strict 50-character subject lines and omitting body text for concise Git commits.

- Tags: how-to-guide
- Published: 2026-07-06

### [How to Install Caveman for Claude Code, Cursor, or Other AI Agents](/JuliusBrussee/caveman/how-to-install-caveman-for-claude-code-cursor-or-other-agents)

Easily install Caveman for Claude Code, Cursor, and AI agents with a single command. Auto-detects and configures plugins for seamless integration.

- Tags: getting-started
- Published: 2026-07-06

### [Caveman Intensity Levels Explained: When to Use Lite, Full, Ultra, and Wenyan Modes](/JuliusBrussee/caveman/what-are-the-different-caveman-intensity-levels-and-when-to-use-each)

Understand Caveman intensity levels Lite, Full, Ultra, and Wenyan. Learn when to use each mode to optimize token usage while preserving technical accuracy and readability for your projects.

- Tags: deep-dive
- Published: 2026-07-06

### [How Caveman Mode Reduces Tokens by 65% While Preserving Technical Accuracy in JuliusBrussee/caveman](/JuliusBrussee/caveman/how-does-caveman-mode-reduce-tokens-while-maintaining-technical-accuracy)

Discover how caveman mode in JuliusBrussee/caveman cuts tokens by 65%. Learn about prompt rules and middleware that preserve accuracy while reducing costs.

- Tags: deep-dive
- Published: 2026-07-06

### [How to Troubleshoot Common Caveman Errors: A Complete Diagnostic Guide](/JuliusBrussee/caveman/how-to-troubleshoot-common-caveman-errors)

Fix common Caveman errors by diagnosing flag-file permissions, stdin handling, and corrupted history files. This guide offers solutions for seamless operation.

- Tags: how-to-guide
- Published: 2026-07-05

### [What Are the Build Targets in the Caveman Makefile?](/JuliusBrussee/caveman/what-are-the-build-targets-in-the-caveman-makefile)

Discover the build targets in the Caveman Makefile: all, build, and clean. Understand compilation, Go builds, and artifact removal for efficient development.

- Tags: how-to-guide
- Published: 2026-07-04

### [How to Update Caveman to the Latest Version](/JuliusBrussee/caveman/how-to-update-caveman-to-the-latest-version)

Easily update Caveman to the latest version by re-running the official installer. It refreshes AI agent integrations with the newest code from the main branch.

- Tags: how-to-guide
- Published: 2026-07-04

### [Where to Find Documentation for Caveman: Complete Repository Guide](/JuliusBrussee/caveman/where-to-find-documentation-for-caveman)

Easily find Caveman documentation within the JuliusBrussee/caveman GitHub repository. Access installation guides, command references, and integration files from the main README.

- Tags: getting-started
- Published: 2026-07-04

### [What is the License for Caveman? MIT License Explained](/JuliusBrussee/caveman/what-is-the-license-for-caveman)

Discover the MIT License for Caveman. Learn how this permissive license allows free use, modification, and distribution of the Caveman software. Understand your rights and obligations.

- Tags: api-reference
- Published: 2026-07-04

### [What Kind of Projects Can Be Built with Caveman: A Complete Guide to Token-Efficient AI Development](/JuliusBrussee/caveman/what-kind-of-projects-can-be-built-with-caveman)

Build custom assistants, microservices, and memory-efficient bots with Caveman. This token-efficient AI development tool reduces output by 65% while preserving code.

- Tags: getting-started
- Published: 2026-07-04

### [How to Test Caveman Locally: Complete Verification Guide](/JuliusBrussee/caveman/how-to-test-caveman-locally)

Learn how to test Caveman locally with our complete verification guide. Run a Python script to validate skills, files, manifests, syntax, scripts, and the Claude hook install flow.

- Tags: how-to-guide
- Published: 2026-07-04

### [Is Caveman Cross-Platform Compatible? macOS, Linux, and Windows Support Explained](/JuliusBrussee/caveman/is-caveman-cross-platform-compatible)

Is caveman cross-platform compatible? Discover how Caveman supports macOS, Linux, and Windows with its unified Node.js installer for seamless development across all major operating systems.

- Tags: getting-started
- Published: 2026-07-04

### [Main Features of Caveman: Token-Saving Skill for AI Coding Agents](/JuliusBrussee/caveman/what-are-the-main-features-of-caveman)

Discover Caveman, the AI coding agent skill that cuts output tokens by 65%. Learn about its compression, developer commands, and middleware features to save AI costs.

- Tags: main-features
- Published: 2026-07-04

### [How to Set Up a Development Environment for Caveman: Complete Guide](/JuliusBrussee/caveman/how-to-set-up-a-development-environment-for-caveman)

Set up your Caveman development environment easily. Clone the JuliusBrussee/caveman repo, install Node.js, and deploy hooks and rule files.

- Tags: how-to-guide
- Published: 2026-07-04

### [Caveman Project Structure: Inside the Node-Based CLI Architecture](/JuliusBrussee/caveman/how-is-the-caveman-project-structured)

Explore the Caveman project structure, a four-layer Node.js architecture featuring a core CLI runtime, modular skills, and hooks for efficient mode management. Understand its design.

- Tags: architecture
- Published: 2026-07-04

### [What Does the `make release` Command Do in Caveman?](/JuliusBrussee/caveman/what-does-the-make-release-command-do-in-caveman)

Automate your Caveman skill release with 'make release'. This command handles versioning, building, checksums, Git tagging, and CI triggers for npm and GitHub Releases.

- Tags: how-to-guide
- Published: 2026-07-04

### [How to Use the caveman Build Command: 4 Intensity Levels Explained](/JuliusBrussee/caveman/how-to-use-the-caveman-build-command)

Master the caveman build command and its four intensity levels lite full ultra and wenyan Achieve faster builds and better code quality with JuliusBrussee/caveman

- Tags: how-to-guide
- Published: 2026-07-04

### [What Is the Purpose of the caveman CLI? Token Compression for LLM Coding Assistants](/JuliusBrussee/caveman/what-is-the-purpose-of-the-caveman-cli)

Discover the caveman CLI, a tool that compresses LLM output by 65% for coding assistants, preserving essential code and text

- Tags: how-to-guide
- Published: 2026-07-04

### [How to Contribute to the Caveman Project: A Complete Guide for Developers](/JuliusBrussee/caveman/how-to-contribute-to-the-caveman-project)

Learn how to contribute to the Caveman project. Follow our guide to edit files, run tests, and submit pull requests using Conventional Commits. Start contributing today

- Tags: how-to-guide
- Published: 2026-07-04

### [What Programming Language Is Caveman Written In? A Technical Analysis](/JuliusBrussee/caveman/what-programming-language-is-caveman-written-in)

Discover the programming language behind Caveman. Primarily JavaScript on Node.js with a Python component for benchmarking. Learn more about this technical analysis.

- Tags: technical-analysis
- Published: 2026-07-04

### [How to Run Caveman Locally: Complete Setup Guide for AI Coding Agents](/JuliusBrussee/caveman/how-to-run-caveman-locally)

Learn how to run Caveman locally with this complete setup guide. Install this token-shrinking AI coding agent skill using a single Node.js script and auto-detect supported agents.

- Tags: how-to-guide
- Published: 2026-07-04

### [What Are the Dependencies for Caveman? Zero-External-Dependency Analysis](/JuliusBrussee/caveman/what-are-the-dependencies-for-caveman)

Discover Caveman zero-external-dependency analysis. This Node.js tool needs only Node.js 18+ and the standard library. Simplify your project with zero npm dependencies.

- Tags: how-to-guide
- Published: 2026-07-04

### [How to Install JuliusBrussee/caveman: One-Liner, Dry-Run, and Manual Methods](/JuliusBrussee/caveman/how-to-install-juliusbrussee-caveman)

Easily install JuliusBrussee/caveman using one-liners, dry-run previews, or manual setup. Get your AI agents up and running quickly and securely.

- Tags: how-to-guide
- Published: 2026-07-04

### [How the Caveman Project Structures Its Skills: A Modular Guide to SKILL.md and README.md Contracts](/JuliusBrussee/caveman/how-does-caveman-project-structure-its-different-skills)

Discover how the Caveman project structures skills using SKILL.md and README.md for modularity and automatic registration. Learn about clear contracts and documentation.

- Tags: architecture
- Published: 2026-07-03

### [Key Insights from Reading Caveman Repository Files: Architecture and Security Analysis](/JuliusBrussee/caveman/what-insights-can-be-gained-from-reading-key-files-of-caveman-repository)

Explore the caveman repository's architecture and security. Discover how it compresses AI outputs, ensures state persistence, and tracks real-time modes across IDEs for enhanced security.

- Tags: architecture
- Published: 2026-07-03

### [What Is the caveman-commit Skill and How It Relates to Other Caveman Skills](/JuliusBrussee/caveman/what-is-purpose-of-caveman-commit-skill-in-relation-to-other-skills)

Discover the caveman-commit skill and its role in generating compressed Conventional Commit messages. Learn how it complements other caveman compression engine features for efficient version control.

- Tags: how-to-guide
- Published: 2026-07-03

### [Understanding the Primary Function of the Caveman Project](/JuliusBrussee/caveman/how-can-i-understand-primary-function-of-caveman-project)

Discover the primary function of the Caveman project. Learn how it injects a speech rule to cut AI output tokens by 65% with 100% technical accuracy, improving efficiency.

- Tags: getting-started
- Published: 2026-07-03

### [What Are the Dependencies for the Caveman Project?](/JuliusBrussee/caveman/what-are-the-dependencies-for-caveman-project)

Discover the dependencies for the Caveman project. This Node.js project requires only version 18+ and uses zero external npm packages, relying solely on the standard library.

- Tags: architecture
- Published: 2026-07-03

### [Caveman Skills Examples: How to Use the JuliusBrussee/caveman Plugin](/JuliusBrussee/caveman/are-there-any-examples-of-how-to-use-caveman-skills)

Discover caveman skills examples for the JuliusBrussee/caveman plugin like commit review and compress. Reduce AI tokens by 65% with terse communication for accurate code.

- Tags: how-to-guide
- Published: 2026-07-03

### [What Kind of Data Does the Caveman-Compress Skill Handle?](/JuliusBrussee/caveman/what-kind-of-data-does-caveman-compress-skill-handle)

Caveman compress handles natural language memory files like CLAUDE.md and todos.md. It compresses verbose prose while preserving code blocks URLs and shell commands.

- Tags: faq
- Published: 2026-07-03

### [Is the Caveman Project Open Source? License, Repository, and Installation Guide](/JuliusBrussee/caveman/is-caveman-project-open-source)

Discover if the Caveman project is open source. Explore its MIT license details, GitHub repository, and get a straightforward installation guide for easy use and modification.

- Tags: getting-started
- Published: 2026-07-03

### [What Does the Caveman Project Name Mean? Explaining the AI Compression Metaphor](/JuliusBrussee/caveman/what-does-caveman-in-project-name-refer-to)

Discover the caveman project AI compression metaphor. Understand how it achieves terse communication, compressing AI agent output by 65% while keeping code and commands intact.

- Tags: explainer
- Published: 2026-07-03

### [How to Contribute to the JuliusBrussee/caveman Project: Complete Guide](/JuliusBrussee/caveman/can-i-contribute-to-juliusbrussee-caveman-project)

Contribute to the JuliusBrussee/caveman project easily. Learn how to edit source files, run tests locally, and adhere to Conventional Commits for a smooth contribution process.

- Tags: how-to-guide
- Published: 2026-07-03

### [Performance Characteristics of the caveman-compress Skill: O(n) Linear Text Compression](/JuliusBrussee/caveman/what-are-the-performance-characteristics-of-caveman-compress-skill)

Explore caveman-compress skill's O(n) linear performance. Achieve 15-30% token reduction with a single-pass regex algorithm preserving code, URLs, and identifiers.

- Tags: performance
- Published: 2026-07-03

### [caveman-commit Skill Use Cases: Automating Conventional Commits in JuliusBrussee/caveman](/JuliusBrussee/caveman/what-are-the-use-cases-for-caveman-commit-skill)

Explore caveman-commit skill use cases for automating conventional commits. Standardize team workflows, draft commits rapidly, and streamline CI/CD with this powerful Git utility.

- Tags: how-to-guide
- Published: 2026-07-03

### [How to Install and Set Up the Caveman Project for AI Coding Agents](/JuliusBrussee/caveman/how-do-i-install-or-set-up-caveman-project)

Install and set up the Caveman project with a single command. It auto-detects and configures agents like Claude Code, Gemini, and Cursor for AI coding.

- Tags: getting-started
- Published: 2026-07-03

### [What Programming Language Is the Caveman Project Written In?](/JuliusBrussee/caveman/what-programming-language-is-caveman-project-written-in)

Discover the programming language behind the Caveman project. Learn that this Node.js application is built with JavaScript and requires version 18+ to run.

- Tags: getting-started
- Published: 2026-07-03

### [Caveman Project Skills Documentation: Complete Guide to Claude Agent Modes](/JuliusBrussee/caveman/is-there-documentation-for-caveman-projects-skills)

Explore the caveman project's skills documentation with our complete guide to Claude agent modes. Discover triggers, rules, and technical boundaries for efficient Claude interactions.

- Tags: documentation
- Published: 2026-07-03

### [How to Get Detailed Feature Information for the caveman-compress Skill](/JuliusBrussee/caveman/how-can-i-get-more-feature-details-on-caveman-compress-skill)

Discover detailed feature information for the caveman-compress skill. Learn how this skill rewrites memory files into a terse caveman style, reducing tokens by 46% via a validation-backed pipeline.

- Tags: how-to-guide
- Published: 2026-07-03

### [Key Files to Understand the Caveman Project: Architecture and Source Guide](/JuliusBrussee/caveman/what-are-some-key-files-to-understand-caveman-project)

Explore key Caveman project files like README.md and caveman-init.js to quickly grasp its architecture and source code. Dive into the token-compression skill today.

- Tags: architecture
- Published: 2026-07-03

### [How the Caveman Project Main Skill Is Implemented: A Technical Deep Dive](/JuliusBrussee/caveman/how-is-caveman-projects-main-skill-implemented)

Uncover how the caveman project main skill is implemented. This technical deep dive explores markdown rules, Node.js activation hooks, and stateless skill architecture for Claude.

- Tags: deep-dive
- Published: 2026-07-03

### [What Are the Other Skills Besides the Main Caveman Skill?](/JuliusBrussee/caveman/what-are-the-other-skills-besides-the-main-caveman-skill)

Discover eight auxiliary skills in the caveman project, including caveman-help, caveman-commit, and more. Extend core compression with helpers for commits, reviews, and stats.

- Tags: getting-started
- Published: 2026-07-03

### [Caveman Project Implementation Details: Architecture, Hooks, and Runtime Behavior](/JuliusBrussee/caveman/can-i-find-implementation-details-for-caveman-project)

Explore JuliusBrussee/caveman implementation details, covering architecture, hooks, and runtime behavior. Discover how this project uses rule files, hooks, and plugins for AI compression.

- Tags: architecture
- Published: 2026-07-03

### [Caveman-Compress Skill Features: Safe Documentation Compression for Developers](/JuliusBrussee/caveman/what-are-the-features-of-caveman-compress-skill)

Discover Caveman-Compress skill features for safe documentation compression. This tool converts prose to caveman-speak, preserving code blocks and URLs for developers.

- Tags: getting-started
- Published: 2026-07-03

### [How the caveman-commit Skill Works in JuliusBrussee/caveman](/JuliusBrussee/caveman/how-does-the-caveman-commit-skill-work)

Discover how the caveman-commit skill generates terse Conventional Commit messages using LLMs and strict formatting rules from SKILL.md without Git operations.

- Tags: internals
- Published: 2026-07-03

### [What Are the Different Skills Available in the JuliusBrussee/caveman Project?](/JuliusBrussee/caveman/what-are-the-different-skills-in-juliusbrussee-caveman-project)

Explore the seven Claude Code skills in the JuliusBrussee/caveman project. Discover features like conversational compression, session analytics, and automated code review to reduce token usage.

- Tags: how-to-guide
- Published: 2026-07-03