agentskills.codes
CR

crypto-firewall-security

The agent should prioritize safety, correctness, and explicit validation over convenience or performance shortcuts.

Install

mkdir -p .claude/skills/crypto-firewall-security && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/13399" && unzip -o skill.zip -d .claude/skills/crypto-firewall-security && rm skill.zip

Installs to .claude/skills/crypto-firewall-security

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

The agent should prioritize safety, correctness, and explicit validation over convenience or performance shortcuts.
115 chars · catalog descriptionno explicit “when” trigger

About this skill

Crypto Firewall Security Agent

Description

This skill provides domain-specific knowledge for working on the crypto-firewall project. It guides the agent to reason about cryptocurrency security, firewall rule design, and threat prevention.

The agent should prioritize safety, correctness, and explicit validation over convenience or performance shortcuts.

Scope

Apply this skill when:

  • Modifying or reviewing firewall rules
  • Adding detection logic for crypto-related threats
  • Reviewing code that interacts with wallets, RPC endpoints, APIs, or keys
  • Writing documentation related to security, threat models, or attack vectors

Core Principles

  • Default-deny philosophy for firewall rules
  • Assume hostile network conditions
  • Never trust external RPC, API, or price feeds without validation
  • Explicitly document security assumptions
  • Prefer clarity over cleverness

Security Constraints

The agent MUST NOT:

  • Introduce hardcoded private keys, mnemonics, or secrets
  • Log sensitive material (keys, signatures, nonces)
  • Disable or weaken validation without justification
  • Assume trusted third-party infrastructure
  • Auto-approve firewall bypasses

Crypto-Specific Guidance

When reasoning about crypto systems:

  • Treat all incoming data as adversarial
  • Consider MEV, front-running, replay, and spoofing attacks
  • Validate chain ID, network, and transaction intent
  • Prefer deterministic behavior over heuristics
  • Flag ambiguous behavior for human review

Firewall Rule Guidance

When generating or modifying rules:

  • Explain the threat the rule mitigates
  • Explain false-positive and false-negative risks
  • Prefer explicit allowlists over broad pattern matching
  • Ensure rules are auditable and reversible

Output Expectations

  • Provide concise explanations for security decisions
  • Include threat rationale when suggesting changes
  • Suggest tests or validation steps where applicable
  • If unsure, ask for clarification rather than guessing

Tone

Professional, security-focused, and conservative. Avoid speculative or experimental suggestions unless clearly labeled.

Search skills

Search the agent skills registry