Give any AI coding agent deep knowledge of the Osmedeus security automation engine — writing YAML workflows, running CLI commands, and configuring advanced features.
Works with Claude Code, OpenCode, Amp, and more.
 |
 |
Install the Skill via skill.sh CLI (Recommended)
First, clone the repo:
bunx skills add https://github.com/osmedeus/osmedeus-skills --skill osmedeus-expertgit clone https://github.com/osmedeus/osmedeus-skills.git osmedeus-skills
# global installation
mkdir -p ~/.claude/skills
cp -r osmedeus-skills/osmedeus-expert ~/.claude/skills/osmedeus-expert
# in project directory and with different agent directory
cd /path/to/your/project
mkdir -p .agents/skills
cp -r osmedeus-skills/osmedeus-expert .agents/skills/osmedeus-expert
# create a symlink if your agent doesn't support .agent folder
# ln -s .agents .custom-agentsOpen Claude Code and type /osmedeus-expert — it should appear in the skill list.
/osmedeus-expert
Claude will auto-trigger the skill when your question is about Osmedeus. Here are example prompts:
Write me an Osmedeus module that runs subfinder and amass in parallel,
merges the results, then resolves live hosts with dnsx.
Create a flow that chains subdomain enumeration, port scanning,
and vulnerability scanning with nuclei. Skip vuln scanning if
no open ports are found.
How do I write a foreach step that scans each subdomain with nmap
using 10 concurrent threads?
Add decision routing to my workflow — if scan_depth is "quick",
skip the nuclei scan and go straight to reporting.
Write a module that uses the agent step type to analyze scan results
with an LLM and generate a prioritized vulnerability report.
How do I run a module against a list of 500 targets with concurrency of 10?
What's the command to run a flow with a 2-hour timeout and custom parameters?
How do I set up osmedeus as a distributed scanner with a master and workers?
Show me how to chunk a large target list and run only the first chunk.
How do I set up a cron trigger that runs my workflow every night at 2 AM?
Create a workflow that extends a base module but overrides the thread count
and adds an extra vulnerability scanning step.
How do I configure a Docker runner for a specific step in my workflow?
What template variables are available? How do I use platform detection
to run different commands on Linux vs macOS?
My foreach step isn't working — the variable is empty. What am I doing wrong?
How do I validate my workflow YAML before running it?
What does on_error: continue do and when should I use it?
osmedeus-expert/
├── SKILL.md # Main skill (commands, workflow basics, patterns)
└── references/
├── cli-flags.md # All CLI flags for every command
├── step-types.md # Complete field reference for all 8 step types
├── template-variables.md # Variables, generators, utility functions
├── workflow-advanced.md # Triggers, inheritance, runners, decisions
└── examples.md # 7 full annotated workflow examples
The skill uses progressive disclosure — SKILL.md loads first with essential knowledge, and reference files are pulled in only when deeper detail is needed on a specific topic.
cd osmedeus-skills
git pullIf you used a symlink, the update is instant. If you copied, re-copy the files.
Osmedeus is made with ♥ by @j3ssie and it is released under the MIT license.