Skip to content

armgit / mso

Public Imported from GitHub
0

Browser-based graphical shell and control plane for a Linux server you own — real terminal, file manager, live metrics and BYOK AI in one mobile-first workspace. Next.js 16 · React 19 · node-pty · Tailwind v4 · shadcn/ui. Public alpha, self-hosted.

fix(security): require executed passive scan evidence

Rahmanb5bda8c94bf0
diff --git a/.github/workflows/dast.yml b/.github/workflows/dast.yml
index 22667599..a3f356c3 100644
--- a/.github/workflows/dast.yml
+++ b/.github/workflows/dast.yml
@@ -23,8 +23,17 @@ jobs:
with:
persist-credentials: false
+ - name: Setup Node
+ uses: actions/setup-node@820762786026740c76f36085b0efc47a31fe5020 # v7.0.0
+ with:
+ node-version: 22
+
+ - name: Require an explicit passive scan target
+ run: node scripts/check-dast-target.mjs
+ env:
+ MSO_DAST_URL: ${{ vars.MSO_DAST_URL }}
+
- name: Run passive ZAP baseline against configured live instance
- if: ${{ vars.MSO_DAST_URL != '' }}
uses: zaproxy/action-baseline@de8ad967d3548d44ef623df22cf95c3b0baf8b25 # v0.15.0
with:
target: ${{ vars.MSO_DAST_URL }}
diff --git a/docs/CHANGELOG.md b/docs/CHANGELOG.md
index 68441db4..36a9cfb9 100644
--- a/docs/CHANGELOG.md
+++ b/docs/CHANGELOG.md
@@ -12,6 +12,7 @@ this is the *what*, and it is what Settings → About shows as “What's new”.
**Fixed**
+- `security` require executed passive scan evidence
- `browser` retain private viewer connectivity without server access
- `release` align Node support and read-only browser reconnects
- `extensions` lock project controls until initial inspection
diff --git a/docs/PROGRESS.md b/docs/PROGRESS.md
index dab0f247..371e20b4 100644
--- a/docs/PROGRESS.md
+++ b/docs/PROGRESS.md
@@ -1,3 +1,14 @@
+## 2026-09-18 — Require executed passive security scan evidence
+
+The security closeout found that the scheduled hosted ZAP workflow could report
+success while its only scan step was skipped when no target was configured. Add a
+dependency-free target guard before an unconditional passive scan. Missing, unsafe
+or credential-bearing targets fail without echoing their contents; existing scanner
+pins, passive scope, rule classifications and alert failures remain unchanged.
+Tests exercise the CLI refusal and workflow wiring. Security assurance guidance now
+separates successful code/dependency/secret checks from historical alert dismissals
+and passive-scan execution; no dismissal or badge attestation is performed.
+
## 2026-09-18 — Keep private viewer routing client-verifiable
The PR review reproduced a deployment-compatibility regression: a hostname resolving
diff --git a/docs/SECURITY-ASSURANCE.md b/docs/SECURITY-ASSURANCE.md
index e955b747..c705057b 100644
--- a/docs/SECURITY-ASSURANCE.md
+++ b/docs/SECURITY-ASSURANCE.md
@@ -36,6 +36,16 @@ API denial, malformed evidence, analysis failure, branch movement and open findi
fail closed. The CLI remains immediate unless `--wait-seconds` is explicitly supplied,
and `--commit` prevents a newer branch tip from being reported as the requested revision.
+## Passive baseline execution evidence
+
+The hosted baseline requires the repository variable `MSO_DAST_URL` to identify an
+explicit approved HTTPS instance without credentials, query parameters or fragments.
+A missing target fails before ZAP starts; it must never produce a green workflow by
+skipping the scanner. An accepted target alone is not a pass: inspect the completed
+ZAP step, its findings and the report artifact. The existing rule classifications and
+warning/failure behavior remain unchanged. Baseline coverage is unauthenticated and
+passive, not an authenticated penetration test or proof about every host service.
+
## GitHub security controls
The public repository enables:
@@ -92,12 +102,12 @@ Automation cannot establish every ASVS requirement. Deployment architecture, hos
## Current OpenSSF posture gaps
-OpenSSF Scorecard is deliberately treated as external posture evidence rather than a scanner whose score MSO can self-declare. Two current findings remain open and must not be hidden or dismissed just to make the security tab look green:
+OpenSSF Scorecard is deliberately treated as external posture evidence rather than a scanner whose score MSO can self-declare. Two repository-process gaps require separate disclosure even when the open-alert count is zero. The September 18 audit confirmed that the corresponding historical alerts were already dismissed by the maintainer on August 31; dismissal is not evidence that the gap was fixed:
-- **Code-Review — High.** The repository currently has one human collaborator, so recent changes do not have independent human approvals. Scorecard explicitly does not count bot/AI reviews as human code review. This closes only after a second human maintainer participates in real review history; enabling a one-approval rule without a second reviewer would only lock the repository and would not create legitimate review evidence.
-- **CII / OpenSSF Best Practices — Low.** The project is not yet enrolled in the OpenSSF Best Practices program. Enrollment and its questionnaire are maintainer-attested work and must not be faked by adding a badge image without registering the project.
+- **Code-Review — High, historical alert #64 dismissed as `won't fix`.** The recorded maintainer rationale is a solo-maintainer workflow without a second active reviewer. Scorecard explicitly does not count bot/AI reviews as human code review. This closes only after a second human maintainer participates in real review history; enabling a one-approval rule without a second reviewer would only lock the repository and would not create legitimate review evidence.
+- **CII / OpenSSF Best Practices — Low, historical alert #65 dismissed as `won't fix`.** The recorded rationale excludes badge enrollment from the release scope. Enrollment and its questionnaire are maintainer-attested work and must not be faked by adding a badge image without registering the project.
-These are repository-process posture findings, not detected application CVEs or leaked secrets. They remain part of the public security evidence.
+These are repository-process posture findings, not detected application CVEs or leaked secrets. They remain part of the public security evidence. The audit also observed thirteen other historical CodeQL/Scorecard dismissals classified as false positives or test usage. Re-read their recorded rationale before relying on them; this audit did not dismiss, delete or reopen any alert. Counts are dated evidence, not a permanent clean-state guarantee.
## What is safe to say publicly
@@ -105,4 +115,4 @@ When the latest `main` workflows and the latest full assurance run are green, th
> MSO is continuously tested with CodeQL, Semgrep, Trivy, OSV-Scanner, Gitleaks, ShellCheck and Dependency Review, with a passive OWASP ZAP production baseline and a separate component-partitioned Codex Security review covering the tracked repository. The public repository has private vulnerability reporting, secret scanning/push protection, dependency alerts, an SBOM workflow and MSO-specific security regression tests. OpenSSF Scorecard is published separately as repository-process posture evidence, and any open Scorecard findings are disclosed rather than folded into a zero-finding claim.
-Do **not** claim “no unresolved High/Critical findings across all gates” while a High Scorecard posture finding is open. Also do **not** shorten the evidence into “certified secure”, “OWASP certified”, “penetration-tested by OWASP”, “100% secure”, or “third-party audited”. Those claims are not established by this process.
+Do **not** claim “no unresolved High/Critical findings across all gates” while a High Scorecard posture gap remains unresolved, even when its GitHub alert has been dismissed. Also do **not** shorten the evidence into “certified secure”, “OWASP certified”, “penetration-tested by OWASP”, “100% secure”, or “third-party audited”. Those claims are not established by this process.
diff --git a/scripts/check-dast-target.mjs b/scripts/check-dast-target.mjs
new file mode 100644
index 00000000..2aa04bb9
--- /dev/null
+++ b/scripts/check-dast-target.mjs
@@ -0,0 +1,23 @@
+#!/usr/bin/env node
+// A configured passive scan must execute; an absent target is not security evidence.
+import { pathToFileURL } from "node:url";
+
+export function requireDastTarget(value) {
+ if (typeof value !== "string" || !value.trim()) {
+ throw new Error("MSO_DAST_URL is required. Configure the approved HTTPS instance before running the passive baseline.");
+ }
+ if (value.length > 2048 || value !== value.trim() || /[\u0000-\u0020\u007f]/.test(value)) {
+ throw new Error("MSO_DAST_URL must be a bounded URL without whitespace or control characters.");
+ }
+ let url;
+ try { url = new URL(value); } catch { throw new Error("MSO_DAST_URL must be an absolute HTTPS URL."); }
+ if (url.protocol !== "https:" || !url.hostname || url.username || url.password || url.search || url.hash) {
+ throw new Error("MSO_DAST_URL must use HTTPS without credentials, query parameters or a fragment.");
+ }
+ return url.href;
+}
+
+if (process.argv[1] && import.meta.url === pathToFileURL(process.argv[1]).href) {
+ try { requireDastTarget(process.env.MSO_DAST_URL); console.log("DAST target validated; the passive scan must still complete."); }
+ catch (error) { console.error(error.message); process.exitCode = 1; }
+}
diff --git a/scripts/dast-target.test.ts b/scripts/dast-target.test.ts
new file mode 100644
index 00000000..c8ae7ab3
--- /dev/null
+++ b/scripts/dast-target.test.ts
@@ -0,0 +1,34 @@
+import { describe, expect, it } from "vitest";
+import { readFileSync } from "node:fs";
+import { spawnSync } from "node:child_process";
+import { requireDastTarget } from "./check-dast-target.mjs";
+
+describe("passive DAST requires an explicit safe target", () => {
+ it.each([undefined, "", " ", "not-a-url", "http://scan.example.com", "https://user:[email protected]", "https://scan.example.com/?token=secret", "https://scan.example.com/#secret", "https://scan.example.com\n", "https://scan.example.com/" + "a".repeat(2048)])("rejects an absent or inappropriate target (%#)", value => {
+ expect(() => requireDastTarget(value)).toThrow(/MSO_DAST_URL/);
+ });
+ it("accepts explicit deployment-owned HTTPS roots and paths without making requests", () => {
+ expect(requireDastTarget("https://scan.example.com")).toBe("https://scan.example.com/");
+ expect(requireDastTarget("https://scan.example.com:8443/app")).toBe("https://scan.example.com:8443/app");
+ });
+ it("fails the actual command on missing configuration without echoing secrets", () => {
+ for (const value of ["", "https://user:[email protected]"]) {
+ const result = spawnSync(process.execPath, ["scripts/check-dast-target.mjs"], { encoding: "utf8", env: { ...process.env, MSO_DAST_URL: value } });
+ expect(result.status).toBe(1);
+ expect(result.stderr).toContain("MSO_DAST_URL");
+ expect(result.stderr).not.toContain("private-value");
+ }
+ });
+ it("runs the target guard before an unconditional passive scanner with unchanged alert policy", () => {
+ const workflow = readFileSync(".github/workflows/dast.yml", "utf8");
+ const guard = workflow.indexOf("run: node scripts/check-dast-target.mjs");
+ const scan = workflow.indexOf("uses: zaproxy/action-baseline@");
+ expect(guard).toBeGreaterThan(0); expect(scan).toBeGreaterThan(guard);
+ expect(workflow).not.toMatch(/^\s*if:/m);
+ expect(workflow).toContain("MSO_DAST_URL: ${{ vars.MSO_DAST_URL }}");
+ expect(workflow).toContain("fail_action: true");
+ expect(workflow).toContain("allow_issue_writing: false");
+ expect(workflow).toContain("rules_file_name: security/zap-baseline.conf");
+ expect(workflow).not.toContain("-I");
+ });
+});