xerj
Skill0 starsChanged 2 months agoMITView on GitHub
Local-first AI search engine, AST-aware code indexer, and Model Context Protocol (MCP) server. Use to auto-index codebases, look up symbol definitions with exact file and line coordinates (xerj def), run BM25/vector/hybrid code search (xerj search), manage durable agent memory, and configure xerj MCP.
- Pipes a download into a shell
- Installs packages
1 other project uses a version of this.
Adopted by
Most recently taken up by
- taixuann/development-spacetheir own version, first seen Nov 8, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- taixuann/development-space★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository20 other files it gives its agents.
- AGENTS.mdAGENTS.md
- rules/AGENTS.mdAGENTS.md
- kurokoSkill
- tayoriSkill
- aesthetic-readme-craftSkill
- cli-builderSkill
- codebase-digestSkill
- git-release-craftSkill
- git-repo-craftSkill
- karakuriSkill
- python-clean-codeSkill
- rust-clean-codeSkill
- rust-style-pythonSkill
- systematic-code-verificationSkill
- compose-clean-codeSkill
- bash-clean-codeSkill
- tui-designSkill
- frontend-designSkill
- web-design-guidelinesSkill
- unslopSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 0 are distinctive, and 6 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 50
Run autoindex to sniff and index code, markdown docs, logs, JSONL, and PDFs with AST-aware line tracking:
- Very specific: line 146
When xerj mcp is active, agents gain access to:
- Very specific: line 158
Avoid Dumping Large Files: Use xerj def "<symbol>" or xerj search "<query>" to retrieve specific passages and line coordinates before …
- Very specific: line 159
Auto-Index on Project Entry: Run xerj autoindex . when exploring an unfamiliar or large codebase to prepare the AST index.
- Very specific: line 160
Use Namespace-Isolated Memory: When persisting cross-session knowledge with xerj_memory_store, always set the namespace to the project name …
- Very specific: line 161
Daemon Security: If starting a local background server (xerj --insecure), ensure it only binds to 127.0.0.1.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
xerj Local-First AI Search & MCP Skill
Integrates AI coding assistants with xerj, a local-first search engine, AST-aware code indexer, and Model Context Protocol (MCP) server written in Rust.
1. When to Use
- Token-Efficient Symbol Lookups: Finding function, struct, class, or type definitions across large codebases without reading entire source files (
xerj def "<symbol>"). - Targeted Code & Documentation Search: Executing BM25, vector, or hybrid queries that return ranked passages with exact
file:linecoordinates (xerj search "<query>"). - Codebase Auto-Indexing: Ingesting and structuring codebases using Tree-sitter AST parsing without manual schema configuration (
xerj autoindex <folder>). - Persistent Agent Memory (RAG): Storing and recalling key architectural decisions, session notes, and design patterns across conversational turns using isolated memory namespaces (
xerj_memory_store,xerj_memory_recall). - MCP Client Integration: Bridging local search, symbol resolution, and agent memory into Antigravity, OpenCode, Claude Code, Cursor, and Codex CLI over
stdio(xerj mcp).
2. Installation & Quickstart
Automated Binary Install (Recommended)
Downloads the prebuilt static binary for the detected OS/architecture and installs to ~/.local/bin:
# Linux / macOS
curl -fsSL https://xerj.org/get | shVerify installation:
xerj --versionInstall From Source via Cargo
cargo install --git https://github.com/xerj-org/xerj.git3. Core CLI Workflows
A. Indexing a Project
Run autoindex to sniff and index code, markdown docs, logs, JSONL, and PDFs with AST-aware line tracking:
# Index current workspace
xerj autoindex .
# Inspect the resulting index mappings and schema
xerj autoindex map
# Dry-run extraction check
xerj autoindex . --dry-runB. AST Symbol Definition (xerj def)
Quickly locate where a symbol is defined, its signature, and exact file and line:
xerj def "RustAnalyzer"
xerj def "AuditReport"
xerj def "handle_install"C. Hybrid & Semantic Search (xerj search)
Execute queries across indexed files with ranked results:
# Full-text / hybrid search
xerj search "clippy warning audit"
# Search within specific indexed folders or data types
xerj search "token savings" --limit 5D. Zero-Setup Q&A (xerj ask)
Query codebases or external resources with cited answers:
xerj ask . "How does skill selection work for different stacks?"E. Measure Token Efficiency Gains
xerj gain4. MCP Server Integration
XERJ includes a built-in stdio MCP server that exposes search, symbol lookups, and memory tools directly to agent runtimes.
Automated Setup
# Wires XERJ into detected agent configurations (Claude Code, Cursor, etc.)
xerj initManual Configuration
Antigravity & Gemini CLI (~/.gemini/antigravity.json)
{
"mcpServers": {
"xerj": {
"command": "xerj",
"args": ["mcp"]
}
}
}OpenCode (~/.opencode/mcp.json or project .opencode/mcp.json)
{
"mcpServers": {
"xerj": {
"command": "xerj",
"args": ["mcp"]
}
}
}Claude Desktop & Claude Code (~/.claude.json)
{
"mcpServers": {
"xerj": {
"command": "xerj",
"args": ["mcp"]
}
}
}Available MCP Tools
When xerj mcp is active, agents gain access to:
xerj_search: Full-text and query-DSL search over indexed code and data.xerj_semantic_search: Dense vector kNN search by semantic meaning.xerj_hybrid_search: Combined lexical BM25 and vector ranking.xerj_memory_store: Saves context or notes into durable namespace-isolated indices.xerj_memory_recall: Retrieves stored memories by keyword or semantic similarity.xerj_brain_overview&xerj_brain_link: Knowledge graph relationships and second-brain navigation.
5. Operational Rules for AI Agents
- Avoid Dumping Large Files: Use
xerj def "<symbol>"orxerj search "<query>"to retrieve specific passages and line coordinates before reading files. - Auto-Index on Project Entry: Run
xerj autoindex .when exploring an unfamiliar or large codebase to prepare the AST index. - Use Namespace-Isolated Memory: When persisting cross-session knowledge with
xerj_memory_store, always set the namespace to the project name (e.g.namespace: "karakuri") to prevent memory collisions. - Daemon Security: If starting a local background server (
xerj --insecure), ensure it only binds to127.0.0.1. Never expose--insecureto public networks.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

