How do I install the Mantic Sh skill?

Run npx skillmds add xiaoland/mantic-sh in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Mantic Sh safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Mantic Sh?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Mantic Sh free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Mantic Sh?

xiaoland (@xiaoland) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Mantic Sh by xiaoland — Agent Skill

Name: Mantic Sh
Rating: 4.666666666666667 (3 reviews)
Author: xiaoland

Mantic Sh

Use Mantic.sh for context-aware code search in local repositories. Use when an agent needs intent-based file discovery, symbol definition/reference lookup, impact analysis, or session-aware context retrieval before coding.

xiaoland b9e051f 2.7 KB Updated Sep 10, 2026

File contents

mantic-sh

Simple guidance for deciding when to use Mantic.sh and how to run it effectively.

When to use this skill

Find code by intent when exact keywords are unknown.
Locate likely files before reading a large codebase.
Jump to symbol definitions across packages.
Find symbol references for refactors or code review.
Estimate blast radius with impact-aware search.
Keep retrieval context across a multi-step coding task.

Common trigger phrases:

"where is X implemented"
"find auth/payment/router logic"
"go to definition of X"
"find references of X"
"show impact of changing X"

When not to use Mantic.sh

Exact literal or regex text search (prefer rg).
Mass text operations (for example, scanning all TODO comments).
Interactive fuzzy browsing (prefer fzf style workflows).

Decision flow

If the task is exact text or regex matching, use rg first.
If the task is intent-based discovery, use mantic <query>.
If a symbol name is known, use mantic goto <symbol> or mantic references <symbol>.
If change risk matters, add --impact.
If results are weak, retry with better domain terms or --semantic.

Procedure

Run Mantic without installation:

npx mantic.sh@latest "<query>" --files

Choose output mode:

--files for fast path lists
--json for structured processing
--markdown for readable terminal output

Use symbol intelligence when applicable:

npx mantic.sh@latest goto "<SymbolName>"
npx mantic.sh@latest references "<SymbolName>"

Use deeper analysis only when needed:

npx mantic.sh@latest "<query>" --impact
npx mantic.sh@latest "<query>" --semantic

Restrict scope for precision:

--path <dir>
--code, --test, --config

Open top-ranked files and continue with normal edit/test workflow.

Completion checks

Query type matches the selected search mode.
At least one relevant file or symbol location is found.
Impact was checked before risky changes.
Fallback to rg was used if Mantic output is empty or too broad.

Quick examples

npx mantic.sh@latest "stripe payment webhook" --files
npx mantic.sh@latest goto "UserService"
npx mantic.sh@latest references "handleLogin"
npx mantic.sh@latest "session auth middleware" --impact --json

MCP note

If your environment already exposes Mantic MCP tools (search_files, get_definition, find_references), prefer those tools over shell commands in that environment.

---name: mantic-shdescription: Use Mantic.sh for context-aware code search in local repositories. Use when an agent needs intent-based file discovery, symbol definition/reference lookup, impact analysis, or session-aware context retrieval before coding.---# mantic-shSimple guidance for deciding when to use Mantic.sh and how to run it effectively.## When to use this skill- Find code by intent when exact keywords are unknown.- Locate likely files before reading a large codebase.- Jump to symbol definitions across packages.- Find symbol references for refactors or code review.- Estimate blast radius with impact-aware search.- Keep retrieval context across a multi-step coding task.Common trigger phrases:- "where is X implemented"- "find auth/payment/router logic"- "go to definition of X"- "find references of X"- "show impact of changing X"## When not to use Mantic.sh- Exact literal or regex text search (prefer `rg`).- Mass text operations (for example, scanning all TODO comments).- Interactive fuzzy browsing (prefer `fzf` style workflows).## Decision flow1. If the task is exact text or regex matching, use `rg` first.2. If the task is intent-based discovery, use `mantic <query>`.3. If a symbol name is known, use `mantic goto <symbol>` or `mantic references <symbol>`.4. If change risk matters, add `--impact`.5. If results are weak, retry with better domain terms or `--semantic`.## Procedure1. Run Mantic without installation:```bashnpx mantic.sh@latest "<query>" --files```2. Choose output mode:- `--files` for fast path lists- `--json` for structured processing- `--markdown` for readable terminal output3. Use symbol intelligence when applicable:```bashnpx mantic.sh@latest goto "<SymbolName>"npx mantic.sh@latest references "<SymbolName>"```4. Use deeper analysis only when needed:```bashnpx mantic.sh@latest "<query>" --impactnpx mantic.sh@latest "<query>" --semantic```5. Restrict scope for precision:- `--path <dir>`- `--code`, `--test`, `--config`6. Open top-ranked files and continue with normal edit/test workflow.## Completion checks- Query type matches the selected search mode.- At least one relevant file or symbol location is found.- Impact was checked before risky changes.- Fallback to `rg` was used if Mantic output is empty or too broad.## Quick examples```bashnpx mantic.sh@latest "stripe payment webhook" --filesnpx mantic.sh@latest goto "UserService"npx mantic.sh@latest references "handleLogin"npx mantic.sh@latest "session auth middleware" --impact --json```## MCP noteIf your environment already exposes Mantic MCP tools (`search_files`, `get_definition`, `find_references`), prefer those tools over shell commands in that environment.

Frequently asked questions