feat(workflow-scanner): E11 part 2 — re-interpolation + auth-bypass + WFL prefix + orchestrator

Closes E11. Three new pieces, plus integration:

1. Re-interpolation detector (Appsmith GHSL-2024-277 stealth pattern).
   The scanner now collects env: bindings (key -> source-expression
   text) by walking parsed events whose parentChain includes 'env',
   then for each `${{ env.<KEY> }}` inside run:, re-injects MEDIUM
   if the binding source matches the 23-field blacklist. This
   catches the pattern where developers apply env-indirection but
   then re-interpolate the env var in run:, which cancels the
   mitigation (template substitution happens before shell parsing).

2. Auth-bypass category (Synacktiv 2023 Dependabot spoofing).
   Detects `if: ${{ github.actor == 'dependabot[bot]' }}` and
   variants. MEDIUM, owasp: 'LLM06' (Excessive Agency). Distinct
   from injection — same expression syntax, different threat class.
   Recommendation steers users to `github.event.pull_request.user.login`.

3. severity.mjs OWASP map registration. WFL prefix added to all
   four maps:
   - OWASP_MAP['WFL'] = ['LLM02', 'LLM06']
   - OWASP_AGENTIC_MAP['WFL'] = ['ASI04']
   - OWASP_SKILLS_MAP['WFL'] = []
   - OWASP_MCP_MAP['WFL'] = []
   Empty arrays for skills/MCP are explicit, not omitted — keeps
   `Object.keys(OWASP_MAP)` symmetric across maps.

4. scan-orchestrator.mjs registration. workflowScan added between
   supply-chain and toxic-flow (toxic-flow correlates after primaries).
   Verified via integration: orchestrator emits 9 WFL findings on
   tests/fixtures/workflows/.

Bug fix: extractTriggers in workflow-yaml-state.mjs was collecting
sub-properties (`branches:`, `types:`) as triggers. Now tracks the
first nested indent level and ignores anything deeper.

Tests:
- 6 new cases in tests/scanners/workflow-scanner.test.mjs:
  re-interp TP, no-double-count, auth-bypass TP, auth-bypass FP
  (startsWith head_ref is not auth-bypass), OWASP map shape,
  orchestrator import + SCANNERS array entry.
- 2 new fixtures: tp-reinterpolation.yml, auth-bypass-dependabot.yml.
- Existing 14 scanner tests + 15 state-machine tests unchanged.

Test count: 1732 -> 1738 (+6). Wave B total: +53 over baseline 1685.
Pre-compact-scan flake unchanged (passes in isolation).
This commit is contained in:
Kjell Tore Guttormsen 2026-04-30 15:57:10 +02:00
commit ede37219a3
7 changed files with 180 additions and 4 deletions

View file

@ -116,6 +116,10 @@ const SINK_PARENTS = new Set(['run']);
// engine, NOT the shell. These are sink mismatches, not injection.
const NON_SINK_PARENTS = new Set(['if', 'with', 'env', 'name', 'runs-on', 'timeout-minutes', 'continue-on-error']);
// B4: auth-bypass — github.actor or forgejo.actor compared against a
// bot identity in if: contexts (Synacktiv 2023 Dependabot spoofing).
const AUTH_BYPASS_RE = /\b(?:github|forgejo)\.actor\s*(?:==|!=)\s*['"][\w-]+\[bot\]['"]/;
// ---------------------------------------------------------------------------
// Discovery
// ---------------------------------------------------------------------------
@ -241,19 +245,85 @@ async function scanFile(absPath, targetPath, stderrLog) {
);
}
const platformLabel = platform === 'forgejo' ? 'Forgejo' : 'GitHub';
const triggerList = [...triggers].join(', ') || 'unknown';
// B4: collect env: bindings (key -> source-expression). Used for
// re-interpolation detection. A binding is an event whose parent is
// a key under an `env:` block — i.e. parentChain includes 'env'
// and the parent is not 'env' itself.
const envBindings = new Map();
for (const ev of parsed.events) {
if (!ev.parentChain.includes('env')) continue;
if (ev.parent === 'env') continue;
if (SINK_PARENTS.has(ev.parent)) continue;
if (NON_SINK_PARENTS.has(ev.parent)) continue;
envBindings.set(ev.parent, ev.expr);
}
for (const ev of parsed.events) {
// B4: auth-bypass first — fires only on if: events
if (ev.parent === 'if' && AUTH_BYPASS_RE.test(ev.expr)) {
findings.push(finding({
scanner: SCANNER_PREFIX,
severity: SEVERITY.MEDIUM,
title: `Actor auth-bypass: ${platformLabel} workflow trusts bot identity`,
description:
`Actor auth-bypass: if-condition trusts bot identity that can be ` +
`spoofed via pull_request_target. ${platformLabel} workflow at ` +
`${relPath}: ${ev.expr}.`,
file: relPath,
line: ev.line,
evidence: `\${{ ${ev.expr} }}`,
owasp: 'LLM06',
recommendation:
'Use `github.event.pull_request.user.login` (immutable per PR) ' +
'instead of `github.actor` for authorization decisions. The actor ' +
'name can be spoofed via Synacktiv-2023 Dependabot path. If the ' +
'check must remain, gate it on an `id-token` OIDC claim.',
}));
continue;
}
if (NON_SINK_PARENTS.has(ev.parent)) continue;
if (!SINK_PARENTS.has(ev.parent)) continue;
// B4: re-interpolation pattern — `${{ env.<KEY> }}` inside run:
// where <KEY> was bound from a blacklisted field via top-level
// or job-level env:. Cancels the env-indirection mitigation.
const reinterpMatch = ev.expr.match(/^env\.([\w-]+)$/);
if (reinterpMatch) {
const key = reinterpMatch[1];
const source = envBindings.get(key);
if (source && DANGEROUS_RE.test(source)) {
findings.push(finding({
scanner: SCANNER_PREFIX,
severity: SEVERITY.MEDIUM,
title: `Re-interpolation: env.${key} re-injects ${source.split(/\s+/)[0]} at ${platformLabel} run:`,
description:
`Re-interpolation: env.${key} was set from \${{ ${source} }}; reading via ` +
`\${{ env.${key} }} in run: re-injects the unsafe value (Appsmith ` +
`GHSL-2024-277 stealth pattern). Workflow: ${relPath}.`,
file: relPath,
line: ev.line,
evidence: `\${{ env.${key} }}`,
owasp: 'LLM02',
recommendation:
`Consume the env var via \$${key} (shell variable) inside run:, ` +
`not via \${{ env.${key} }}. Template substitution happens before ` +
`shell parsing — re-interpolating cancels the env-indirection ` +
'mitigation and re-introduces the original injection.',
}));
continue;
}
}
const fieldClass = classifyField(ev.expr);
if (fieldClass === 'other') continue;
const severity = severityFor(triggers, fieldClass);
if (!severity) continue;
const platformLabel = platform === 'forgejo' ? 'Forgejo' : 'GitHub';
const triggerList = [...triggers].join(', ') || 'unknown';
findings.push(finding({
scanner: SCANNER_PREFIX,
severity,