Skip to content

Security: wevnal/.github

Security

SECURITY.md

Security Policy

wevnal practices DevSecOps: security is built into our planning, development, review, and operations. We continuously improve our posture through automation, monitoring, and feedback loops across the SDLC.

DevSecOps at wevnal

  • Security is a shared responsibility across product, engineering, and operations.
  • We integrate security into design reviews, code review, automated checks, and runtime observability.
  • We iterate based on findings from security tooling, incidents (if any), and partner/vendor assessments.

Security Assurance Responsibilities

  • Infrastructure Security — Wiz
    We use Wiz to continuously assess cloud posture (CSPM/CNAPP), detect misconfigurations, and prioritize remediation across our infrastructure.
  • Code/Application Security — Aikido
    We use Aikido to identify and manage code-level risks, including dependency vulnerabilities, secret exposure, and common misconfigurations early in development.

These tools complement defense-in-depth practices such as least privilege, secure configuration baselines, secrets management, and monitoring.

Reporting a Vulnerability

Please report suspected vulnerabilities responsibly:

  • Prefer a private report via GitHub Security Advisories for this organization/repo.
  • Or contact us via the inquiry channels on our corporate site: https://wevnal.io/contact

Include:

  • Description of the issue, affected component(s), and potential impact
  • Steps to reproduce or a minimal PoC
  • Relevant logs or screenshots when possible

We will acknowledge receipt, investigate, and work with you to validate and remediate as appropriate. Issues affecting user data or production systems are prioritized.

Scope

  • This policy covers code and projects under the wevnal GitHub organization and the infrastructure used to deliver our services, including BOTCHAN and other properties referenced on https://wevnal.io and https://botchan.chat.
  • Do not perform testing that risks availability, integrity, or privacy (e.g., DDoS, social engineering, or accessing third‑party data without permission).

Responsible Disclosure

We ask that you:

  • Give us reasonable time to remediate before public disclosure
  • Avoid accessing or exfiltrating more data than necessary to demonstrate the issue
  • Follow applicable laws and act in good faith to avoid service disruption or privacy risk

Acknowledgements

We appreciate the security community’s contributions to keeping our ecosystem safe.

There aren’t any published security advisories