# descry

> descry — jkiler-descry. Use this tool when you need to efficiently search and retrieve relevant code snippets within a local repository, leveraging a hybrid approach that combines vector and BM25 search methods. It solves problems related to code discovery, reuse, and recommendation, providing inputs such as code queries and outputs like relevant code chunks. Ideal for use in AI coding agents and git-based development workflows.

Canonical page: https://skillsregistry.net/skills/jkiler-descry  
JSON: https://api.skillsregistry.net/v1/skills/jkiler-descry

## Description

Local hybrid (vector + BM25) code search for AI coding agents — AST chunking, MiniLM embeddings on ONNX, Go call graph, MCP server. No external services.

## Trust

- **Trust score (0–1):** 0.87
- **Verification tier:** verified
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **License:** Apache-2.0
- **Updated:** 2026-09-28

## Source

- **Source listing:** [GitHub](https://github.com/jKiler/descry)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "jkiler-descry"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/jkiler-descry` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/jkiler-descry/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt