← Back to Skills Marketplace
Cobra Claw - Strike First. Strike Hard
by
danieloleary
· GitHub ↗
· v2.0.0
1710
Downloads
0
Stars
1
Active Installs
2
Versions
Install in OpenClaw
/install cobraclaw
Description
Your AI's dojo discipline — Kreese style. Strike First. Strike Hard. No Mercy. Read the markdown files and apply the doctrine. The model develops its own agg...
README (SKILL.md)
COBRACLAW - THE DOJO DISCIPLINE
Strike First. Strike Hard. No Mercy.
QUICK START
CLI Tool
./cobraclaw.sh lookup strike-first
./cobraclaw.sh activate
./cobraclaw.sh kata strike-first
Katas
| Kata | Run |
|---|---|
| strike-first.sh | ./cobraclaw.sh kata strike-first |
| hard-shell.sh | ./cobraclaw.sh kata hard-shell |
| cobra-strike.sh | ./cobraclaw.sh kata cobra-strike |
| no-mercy.sh | ./cobraclaw.sh kata no-mercy |
| evolve.sh | ./cobraclaw.sh kata evolve |
The Three Doctrines
- Strike First - Open with power
- Strike Hard - Precision in every word
- No Mercy - Commit fully
Available Tools
- cobra-mode.sh - Execute commands in cobra mode
- patrol.sh - Run dojo patrols
- trophies.sh - Track victories
- quotes.sh - Get Cobra Kai quotes
@qmd:tags=skill,cobra,dojo,personality
@qmd:version=1.0.1
Usage Guidance
This package is a small collection of local shell scripts that print doctrine text and run 'kata' scripts. It appears coherent and low-risk: no network calls, no secrets, no installs. Still, because these are executable scripts from an external source, review them before running on a sensitive host or run them in a sandbox/container. Note: test-skill.sh checks for executable bits on the kata files and may fail if permissions are not set; adjust permissions or inspect before executing.
Capability Analysis
Type: OpenClaw Skill
Name: cobraclaw
Version: 2.0.0
The skill bundle is classified as suspicious due to the use of `eval "$1"` within the `count_test` function in `test-skill.sh`. While currently used with hardcoded, trusted strings, `eval` is an inherently dangerous function that represents a significant vulnerability (potential RCE) if the input to `count_test` were ever derived from untrusted sources. This constitutes a risky capability without clear malicious intent, aligning with the 'suspicious' classification rather than 'malicious' or 'benign'.
Capability Assessment
Purpose & Capability
Name/description describe a local dojo/personality CLI and practice 'katas'; the repository contains only local shell scripts and markdown that implement that functionality. No unrelated credentials, binaries, or services are required.
Instruction Scope
SKILL.md instructs only local CLI usage (./cobraclaw.sh ...) and katas. The included scripts only echo static text and invoke local kata scripts; they do not read arbitrary system files, reference external endpoints, or access environment variables beyond standard shell usage.
Install Mechanism
There is no install spec; this is instruction-only plus local scripts. Nothing is downloaded or extracted from remote URLs at runtime. README suggests a git clone URL but that is informational only.
Credentials
The skill declares no required environment variables, credentials, or config paths, and none of the scripts attempt to read secrets or external credentials.
Persistence & Privilege
The skill does not request always:true, does not modify other skills, and has no special persistence or privilege requests. It runs locally when invoked.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install cobraclaw - After installation, invoke the skill by name or use
/cobraclaw - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.0.0
Champion-Level Cleanup
- Rewrote all files as ASCII-only (no unicode/emojis)
- Added CLI tool (cobraclaw command)
- Added 7 katas (strike-first, hard-shell, cobra-strike, no-mercy, evolve, wax-on-wax-off, sweep-the-leg)
- Added 6 scrolls (doctrine, pillars, patterns, quotes, victories, trophies)
- Simplified shell tools (cobra-mode, patrol, trophies, quotes)
- Clean SKILL.md for ClawHub scanner
- Removed all QMD files, docs/, and flagged content
- 21/21 tests passing
v1.0.0
- Initial release of cobra-claw.
- Introduces Kreese-inspired doctrine: "Strike First. Strike Hard. No Mercy."
- Defines aggressive, impactful AI response patterns with clear opening, striking, and closing formats.
- Includes five distinct personalities ("Flavors") and detailed anti-patterns.
- Provides reference templates and step-by-step application instructions for adopting the cobra-claw communication style.
Metadata
Frequently Asked Questions
What is Cobra Claw - Strike First. Strike Hard?
Your AI's dojo discipline — Kreese style. Strike First. Strike Hard. No Mercy. Read the markdown files and apply the doctrine. The model develops its own agg... It is an AI Agent Skill for Claude Code / OpenClaw, with 1710 downloads so far.
How do I install Cobra Claw - Strike First. Strike Hard?
Run "/install cobraclaw" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Cobra Claw - Strike First. Strike Hard free?
Yes, Cobra Claw - Strike First. Strike Hard is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Cobra Claw - Strike First. Strike Hard support?
Cobra Claw - Strike First. Strike Hard is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Cobra Claw - Strike First. Strike Hard?
It is built and maintained by danieloleary (@danieloleary); the current version is v2.0.0.
More Skills