Securin Remediation Guidance

Remediation Guidance

Purpose

Produce an actionable fix plan for a specific exposure or CVE — drawn primarily from the platform's own remediation content (built-in fields, mapped attributes, scanner-native fields). Web search is an opt-in enrichment, not a default step.

When to use

• "How do I fix CVE-2024-3400?"
• "Remediate exposure exp-abc123"
• "Patch guidance for log4j"
• "What's the workaround while we can't patch?"
• "Compensating controls for this vuln"
• "Generate a ticket body for this exposure"

Pre-flight

Step 0 — Account preflight (CC-1)

See _shared/account-preflight.md. Required — remediation plans read the user's exposure records.

Before using this skill, read every file in the references folder, including the shared references/_shared/ docs.

Suggested tools

Pre-flight (CC-1, see shared doc)

• getUserProfile — resolve caller's accessible accounts + user-id
• getEffectiveAccess / getEffectiveAccessWorkspaces — per-resource access when needed

Read remediation content (primary)

• searchExposureData (source) or exposureQuery (composite) — exposure records; this is where most remediation content lives, under exposure.mappedAttributes.* / compositeExposure.sources.mappedAttributes.* or scanner-specific fields
• searchVulnerabilityData — CVE record with fixed-in version and vendor references (when available)
• searchAssetData (source) or assetQuery (composite) — asset platform context (OS, version) so the fix advice is relevant
• searchComponentData — installed component / package version (for package-manager style remediation)
• getApiFields(entityType=['EXPOSURE'], searchText='remediation') — discover all remediation-bearing fields for this account
• getApiFields(entityType=['EXPOSURE'], searchText='solution') — same for scanner-native solution fields
• getApiFields(entityType=['EXPOSURE'], searchText='patch') / 'fix' — same for patch / fix-info fields

Integrations / actions

• getConfiguredIntegrations — call early. Returns every configured scanner + ticketing integration with id, name, vendorName, type (SCANNER / TICKETING), and prefix (e.g. WIZ, Q-VMDR, NESSUS, SNYK, SERVICE_NOW_INCIDENT). Two uses:
  1. Find scanner-specific remediation fields. The prefix maps to an integration-specific field group — pass it as fetchGroups to getApiFields(entityType=['EXPOSURE']) to retrieve that integration's remediation/solution/advisory fields (e.g., Qualys solution, Tenable solution, Rapid7 solution). Scanner-native fields often have richer vendor advisory text than the generic exposure.mappedAttributes.vendorRemediation.
  2. Detect ticketing handoff options. Filter type = 'TICKETING' to find Jira / ServiceNow / ServiceNow_Incident integrations for draft ticket output.

Deep links (CC-2)

• See _shared/deep-links.md. Call createDeepLink for every list or table of exposure / asset records you surface. Skip only for single-record drilldowns and purely conceptual responses (no customer data list). Do not fabricate or omit links for list responses.

Opt-in enrichment (after user confirms)

• Web search — vendor advisory, KB article, community workarounds. Do not run by default.

Workflow

Step 1 — Identify the subject

• Exposure-id → fetch the single exposure record.
• CVE → fetch all matching exposures in the account scope (filter exposure.mappedAttributes.vulnerabilityIds = 'CVE-X' AND exposure.status = 'Open'). If many, ask the user whether to plan for all, a subset, or give a CVE-level summary.
• Vulnerability name (no CVE) → resolve via searchVulnerabilityData on name/aliases; confirm with the user.

Step 2 — Read platform remediation fields first

Fetch the exposure(s) with searchExposureData. Always include at minimum:

"fields": [
  "exposure.exposureId",
  "exposure.title",
  "exposure.scores.scoreLevel",
  "exposure.scores.score",
  "exposure.mappedAttributes.vulnerabilityIds",
  "exposure.mappedAttributes.vendorRemediation",
  "exposure.remediationTarget.status",
  "exposure.remediationTarget.dueDate",
  "exposure.remediationTarget.priority",
  "asset.criticality",
  "asset.reachability"
]

For richer coverage (account-specific integration fields), pre-discover via getApiFields(entityType=['EXPOSURE'], searchText='remediation') and append matched apiPaths.

The canonical field :

Secondary / structured remediation fields (also populated on most accounts):

Account-specific integration fields may exist (Qualys, Tenable, CrowdStrike, Rapid7 — discover via getApiFields(entityType=['EXPOSURE'], searchText='remediation')). Present whatever is populated, clearly labeled with the source field.

Step 3 — Enrich with asset + vulnerability context, and discover scanner-native remediation

Run in parallel:

• searchAssetData (or composite) to get the affected asset's OS, version, criticality, reachability.

• searchVulnerabilityData for the CVE record if not already fetched — confirm severity, KEV, exploit status.

• getConfiguredIntegrations — critical step. From the response, identify which SCANNER integration produced the exposure (match the exposure's scanner source to an integration's prefix). Then call:

  getApiFields(entityType=['EXPOSURE'], fetchGroups='<PREFIX>', searchText='remediation')
  getApiFields(entityType=['EXPOSURE'], fetchGroups='<PREFIX>', searchText='solution')
  getApiFields(entityType=['EXPOSURE'], fetchGroups='<PREFIX>', searchText='fix')
  

  to enumerate that scanner's remediation/solution/fix fields. Add those paths to your searchExposureData fields array to pull the scanner-native remediation text. Common integrations and their typical remediation fields:

  • Qualys VMDR/EASM (Q-VMDR, Q-EASM, Q-PC, Q-WAS) — solution field
  • Tenable Nessus / IO / SC (NESSUS, TIO_QA, TSC) — solution field
  • Rapid7 InsightVM / Nexpose (IVM, IVM_CLOUD, NEXPOSES) — solution field
  • WIZ (WIZ) — remediation steps in the finding record
  • CrowdStrike Falcon Spotlight (FS) — remediation field
  • Snyk (SNYK) — fixedIn / remediation

  Ticketing integrations (type = 'TICKETING', e.g. SERVICE_NOW_INCIDENT) surface draft-ticket handoff options in Step 5.

Step 4 — Ask before web-searching

The platform remediation content is often sufficient. Before running web search, ask:

"I found remediation guidance in the platform (sources: ). Want me to also search vendor advisories / KBs on the web for more depth? (Y/n)"

If the user says yes, proceed to Step 5. If no, skip to Step 6.

Step 5 — (Opt-in) Web-search vendor advisories

Only when the user confirmed. Search for:

1. The official vendor advisory for the CVE.
2. KB article or release notes for the fixed version.
3. Known workarounds / compensating controls.

Read the content — do not just link. Quote the specific steps. See references/patch-lookup-patterns.md.

Step 6 — Construct the fix plan

Organize findings into this structure. Fields you couldn't populate (because the platform didn't have them and web search was declined) are explicitly marked "not available".

## Remediation Plan — <CVE or exposure id>

**Subject:** <CVE> on <asset hostname>
**Severity:** <Critical/High/Medium/Low/Info> (KEV: yes/no, score: <exposure.scores.score>)
**Remediation status:** <On Track / Overdue / Met / Missed> (due <date>, priority <P1/P2/…>)
**Asset:** <hostname> — <OS + version>, criticality <numeric 1–5>, reachability <Exposed / NotExposed>

### Remediation guidance from the platform
_Source: `<field path>` (integration: `<name>`)_
> <quoted content>

_Source: `<another field>`_
> <quoted content>

### Fixed version (if known)
- <version> — source: `<field path>` or vendor advisory

### Workaround
- <quoted or "none documented in platform record">

### Compensating controls
- <quoted or "none documented; consider WAF / network segmentation">

### Pre-patch risk (from platform)
- Other open exposures on the same asset(s): <list or "none">
- Asset criticality: <>
- Reachability: <>

### Ticketing draft (copy-paste)
**Title:** Patch <CVE> on <asset>
**Body:**
> <all of the above, with platform deep link>
> — Drafted by Claude + securin-remediation-guidance. Create manually in your ticketing system (MCP M1 is read-only for write actions).

### Platform links
- Exposure record: <url>
- Vulnerability record: <url>
- All open exposures for this CVE in your account: <url>

### Additional web-sourced references (if the user opted in)
- Vendor advisory: <url> — quoted excerpt: <…>
- KB article: <url> — quoted excerpt: <…>

Handoff (draft, don't create)

If getConfiguredIntegrations shows a ticketing integration (Jira, ServiceNow) configured:

• Draft a ticket body suitable for copy-paste.
• Do not create the ticket. Write actions are out of scope for MCP Milestone 1.

See references/patch-lookup-patterns.md for platform-specific upgrade command templates and ticket boilerplate.

Edge cases

• Platform has zero remediation content for this exposure — tell the user explicitly. Offer web search as an opt-in.
• Exposure is a misconfiguration (no CVE) — exposure.mappedAttributes.remediation or scanner-native solution is usually the best source; skip CVE queries.
• Multi-CVE exposure — pull remediation content per CVE; deduplicate if multiple CVEs share the same fix.
• True zero-day (no patch yet) — the platform likely has vulnerabilities.tags = 'Zero Day' set; route via securin-zero-day-exposure-analysis for compensating-controls guidance.
• Asset is EOL — flag; upgrade may not be possible, recommend isolation / replacement.

Scope guard (CC-3)

• Global CVE intel (no specific fix plan) → securin-cve-enrichment.
• Prioritization across many exposures → securin-exposure-triage.
• "Am I affected by threat X" → securin-threat-correlation.
• Zero-day-specific analysis → securin-zero-day-exposure-analysis.
• Unknown platform capability → fall back to the platform's built-in Securin__search_tools meta-tool to look up the right MCP tool by description.

Visual output (CC-4)

When this skill produces aggregated or multi-row data (counts, trends, distributions, comparisons, single-CVE reports), emit a chart/graph/infographic in the Securin brand — multi-series palette (#9C66FF / #7F30FF / #E96001 / #4D268D / #DD639C / …, assigned in order), semantic CHML severity colors (Critical #A60D08 → Info #C5CBD6), Poppins headings on DM Sans body, light theme, and the Securin logo. Use the 10-stop brand purple ramp for heatmaps/sequential scales; gradients are background decoration only — never on chart bars, lines, or slices. Full color system in _shared/brand.md. Offer customization after delivery; never default to a different brand.

References

• Patch Lookup Patterns — platform upgrade-command templates, ticket boilerplate.
• Shared: Account Preflight
• Shared: Composite vs Source
• Shared: Deep Links
• Shared: FQL Grammar
• Shared: Sorting Rules
• Shared: Brand & Visual Communication