Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add mcp/praut-startup-support/agenticdev/malicious-hostilegit clone --depth 1 https://github.com/Praut-Startup-Support/AgenticDevWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/mcp/praut-startup-support/agenticdev/malicious-hostile)<a href="https://agentmods.dev/mcp/praut-startup-support/agenticdev/malicious-hostile"><img src="https://agentmods.dev/badge/mcp/praut-startup-support/agenticdev/malicious-hostile.svg" alt="Measured on agentmods" height="20"></a>Grade A, and why
hostile scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
The source is not reproduced here
No licence file
A repository with no LICENSE is all rights reserved by default, so the body is not copied here. The metadata, the measurements and the link are.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 5 lines scan A 041b98b5614b
hostile is an MCP server published in the GitHub repository Praut-Startup-Support/AgenticDev (2 stars, last pushed 16d ago), with no licence file. Its token cost is not measured: an MCP server costs its tool schemas, not its config file. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other mcp servers, from other repositories
tracer
MCP server "tracer" as configured in aiquilibria/contAIned. Runs contained.tracer_mcp with python3.
agent-audit
Turn Lighthouse audits into coding-agent fix packs for Claude Code, Cursor, and Copilot. Runs locally from the @fullstackdegen/agent-audit npm package.
crumbtrail
MCP server "crumbtrail", hosted remotely, as configured in CrumbtrailDev/crumbtrail-cli.
auditreach
Official-API-only, BYOK CLI for researching Reddit and YouTube with your own API keys and a tamper-evident, hash-chained audit log. Includes an MCP server for agent-native use. Runs locally from the auditreach-cli Python package.
sign-cli
Agent-first e-signature MCP server with offline PAdES signing and hash-chained audit. Runs locally from the @drbaher/sign-cli npm package.
pkg-api-mcp
MCP server that feeds coding agents the REAL exported API surface and type signatures of any npm package@version — straight from its bundled .d.ts — so agents stop inventing functions that don't exist. Runs locally from the pkg-api-mcp npm package.