Vex Agent Skill

VEX skill

This skill is the entrypoint for VEX operations in hve-core. It combines the OpenVEX v0.2.0

specification reference with reusable management playbooks for implementing, reviewing, and

validating VEX documents. The normative reference material below remains the authoritative

source for schema, status logic, and public-source guidance.

VEX management playbooks

Detection, drafting, and attestation are workflow-owned automation. This skill supplies the

reusable procedures, mutation rules, and review criteria. The CVE Analyzer subagent performs the per-CVE exploitability analysis that feeds those workflows.

Implement VEX in a target project

Use this playbook when standing up VEX in a target project. Scaffold the VEX document under

security/vex, wire the vex-detect and vex-draft workflows, reference the PR-body scaffold in

assets/pr-body-scaffold.yml, connect the dedicated reusable VEX

attestation workflow for provenance and OpenVEX-over-SBOM attestation, and set CODEOWNERS on the VEX document.

Use references/vex-status-logic.md and the

vex-standards.instructions.md instructions for the detailed rules.

Review and validate VEX

Use this playbook when reviewing drafted VEX statements. Assess the status determination against

the evidence and confidence bands, honor the document mutation and forbidden-transition contract,

and validate the release attestation output. Attestation generation is owned by the dedicated

reusable VEX attestation workflow, not by the reviewer. The forthcoming tested gate module and

tests will live in this skill so the workflow and interactive entry points can share the same

rules.

VEX statuses

| Status | Meaning |

|-----------------------|------------------------------------------------------------------------------------------------------------|

| not_affected | The vulnerability is not exploitable in this product. Requires a justification or impact_statement. |

| affected | The vulnerability is exploitable. Requires an action_statement describing remediation. |

| fixed | The vulnerability was present but has been remediated in this product version. |

| under_investigation | The author is evaluating whether the vulnerability affects this product. Safe default for uncertain cases. |

Justification codes for not_affected

When a statement uses not_affected status, it must include a machine-readable justification:

| Code | Meaning |

|-----------------------------------------------------|--------------------------------------------------------------------|

| component_not_present | The vulnerable component is not included in the product. |

| vulnerable_code_not_present | The component is present but the vulnerable code is not included. |

| vulnerable_code_not_in_execute_path | The vulnerable code is present but cannot be reached at runtime. |

| vulnerable_code_cannot_be_controlled_by_adversary | The code is reachable but an attacker cannot influence the inputs. |

| inline_mitigations_already_exist | Existing controls prevent exploitation of the vulnerability. |

Product identifiers

Products use Package URL (PURL) format

(for example, pkg:npm/@microsoft/[email protected]).

Normative references

• OpenVEX JSON Schema Reference: field definitions, required

versus optional fields, and example documents.

• VEX Status Logic: status determination decision tree,

evidence requirements per status, and forbidden transitions.

• CVE Data Sources: OSV.dev, NVD, and GitHub Advisory Database

API references with licensing posture.

Skill layout

• SKILL.md: this file (skill entrypoint).
• references/: normative reference documents.
• openvex-schema.md: JSON schema reference with field definitions and examples.
• vex-status-logic.md: status determination decision tree and forbidden transitions.
• cve-data-sources.md: CVE data source API references and licensing.

Attribution and licensing

The OpenVEX specification reference content in this skill is derived from the OpenVEX Community

specification and remains attributed to the OpenVEX Community. The reusable VEX management

playbooks and the surrounding guidance in this skill are hve-core-authored content. The skill

frontmatter uses a mixed-attribution metadata set so the upstream specification reference and the

hve-core playbooks are clearly distinguished.

Third-Party Attribution

| Attribute | Value |

|---------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------|

| Specification | OpenVEX Specification v0.2.0 |

| Copyright | © OpenVEX Contributors |

| License | Apache License 2.0 |

| Source | <https://github.com/openvex/spec/blob/main/OPENVEX-SPEC.md> |

| Modifications | Specification restructured into agent-consumable reference documents with added status determination logic, evidence requirements, and CVE data source guidance. |