Skip to content

Security: AIXP-Labs/AISP

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

If you discover a security vulnerability in the AISP protocol specification or reference implementations, please report it responsibly through GitHub's private security advisory channel:

Report a vulnerability on GitHub

This keeps the report private until a fix is released and coordinated disclosure is complete.

Please include:

  • Description of the vulnerability
  • Steps to reproduce (if applicable)
  • Potential impact assessment
  • Suggested fix (if any)

A Note on the Skill Attack Surface

AISP skills carry executable resources (scripts, templates, data) and a discovery script (aisp_list.py). The most relevant classes of issue are:

  • Resource path escape — a resources[].path that traverses outside the skill folder or _shared/ (../).
  • Undeclared resources — files in a skill folder not listed in aisp_contract.resources (unknown surface).
  • Phantom enforcement — a non_negotiable.enforced_by that points to a node/step/mechanism that does not exist, giving a false sense of a hard guarantee.
  • Discovery script side effects — aisp_list.py doing anything beyond scanning folders and (with --json) writing aisp_list.json.
  • Self-declared trust — a skill claiming to be verified / trusted / safe.
  • Remote resources — resources referencing remote URLs without a user confirmation gate.

Response Timeline

  • Acknowledgment: Within 48 hours
  • Initial Assessment: Within 7 days
  • Resolution Plan: Within 14 days

Scope

This security policy covers:

  • The AISP protocol specification (specification/AISP_Protocol.md)
  • The AISP conformance standards (specification/standards/)
  • The AISP proto definition (specification/aisp.proto)
  • The reference discovery script (examples/aisp/aisp_list.py)
  • Official documentation and examples

Out of scope:

  • Third-party AISP implementations and runtimes (report to those projects directly)
  • The AISOP execution language itself (report to the AISOP-Protocol repository)
  • AIAP governance rules — a separate sibling protocol in the AIXP family (report to the AIAP-Protocol repository)
  • SoulBot runtime issues (report to the SoulBot repository)
  • Skills authored by third parties (report to their respective authors / registries)

Coordinated Disclosure

We follow a coordinated disclosure process. Please do not publicly disclose vulnerabilities until a fix has been released and announced.

Align Axiom 0: Human Sovereignty and Wellbeing. AISP — AI Skill Protocol V1.0.0. www.aisp.dev

There aren't any published security advisories