Documentation
All 33 Pix security workflows delivered as MCP prompts. The 17 that run entirely on the server, and the 16 that drive your own machine.
Tools answer one question. Prompts run a whole workflow.
The 33 prompts are the complete Pix skill library, a set of security procedures written and refined by engineers who do this work. They are delivered verbatim from their source, with CI failing the build if the two ever drift.
| Count | Runs | Needs the CLI | |
|---|---|---|---|
| Server-side | 17 | Entirely through MCP tools | No |
| Local-execution | 16 | Your machine, via your agent | Yes |
Sixteen of them read files, run package managers, or shell out to scanners. A Cloudflare Worker has none of those, so they ship as instructions your own agent follows using the Vulnetix CLI. Every one says so in its first lines:
This workflow needs your local machine. It reads or writes files in the repo, runs package managers, or shells out to scanners. The MCP server has no filesystem, so run these steps yourself with the Vulnetix CLI.
That boundary is the reason no code ever leaves your machine through this server.
Run with nothing installed.
| Prompt | Workflow |
|---|---|
vulnetix_vuln | Full CVE triage with repo impact |
vulnetix_exploits | Exploit intelligence and CWSS priority |
vulnetix_exploits_search | Find actively-exploited issues in your ecosystem |
vulnetix_remediation | Structured remediation playbook |
vulnetix_safe_version | Newest version free of known vulnerabilities |
vulnetix_package_search | Pre-add package risk assessment |
vulnetix_dep_add_guard | ALLOW / WARN / BLOCK verdict on a new dependency |
vulnetix_typosquat_check | Typosquat and malicious-package detection |
vulnetix_kev_watch | KEV additions crossed with your dependencies |
vulnetix_soc_triage | Shift-start triage queue, P1–P4 |
vulnetix_threat_feed | Threat-intel digest |
vulnetix_attack_mapping | ATT&CK technique mapping |
vulnetix_ioc_pivot | IOC pivots, optionally as a STIX bundle |
vulnetix_detection_rules | Snort, YARA and Nuclei content |
vulnetix_exploit_test | Generate a validation command |
vulnetix_secure_code_write | Secure-coding coach for the file you are in |
vulnetix_incident_respond | Full SOC playbook for a CVE gone hot |
Need a repository and the CLI.
| Prompt | Workflow |
|---|---|
vulnetix_sast_scan | SAST over changed files, optional Semgrep |
vulnetix_secret_scan | Hardcoded-secret detection, staged or full |
vulnetix_iac_scan | Terraform, OpenTofu and k8s misconfiguration |
vulnetix_container_scan | Dockerfile and image analysis |
vulnetix_license_check | Copyleft conflicts and SPDX output |
vulnetix_sbom_generate | CycloneDX and SPDX, optionally cosign-signed |
vulnetix_eol_check | End-of-life runtimes and dependencies |
vulnetix_fix | Apply a remediation with rollback |
vulnetix_dep_resolve | Resolve a blocked upgrade |
vulnetix_verify_fix | Re-scan and gate on the result |
vulnetix_vex_publish | OpenVEX and CycloneDX attestations |
vulnetix_compliance_report | SBOM, SARIF, VEX and signatures in one bundle |
vulnetix_code_review_security | Unified pre-merge review of a PR diff |
vulnetix_dashboard | Read tracked findings from local memory |
vulnetix_capabilities_detect | Probe installed binaries and repo signals |
vulnetix_get_api_key | Walk through obtaining a credential |
How prompts surface depends on the client:
| Client | How |
|---|---|
| Claude Code | /mcp__vulnetix__vulnetix_vuln |
| VS Code | /mcp.vulnetix.vulnetix_vuln |
| Claude Desktop | + in the message box, under the server |
| Cursor, Windsurf | Ask by name, or pick from the prompt list |
Most take one free-form argument: an advisory id, a package name, a flag string.
Bodies are generated from the Pix SKILL.md sources by a build script, and CI
fails if the generated file drifts. That is on purpose: the plugin’s own
_system.md and README.md are already stale against its current version, and
documentation that quietly disagrees with the thing it documents is worse than
none.