secops-triage · google/mcp-security — Claude Code skill: triage and install
The secops-triage skill investigates security alerts, checks related cases and SIEM context, enriches indicators, then recommends closure or escalation.
Apache-2.0 21 installs ★ 546 updated 2026-10-27
npx skills add google/mcp-security --skill secops-triage -a claude-codeWhat secops-triage does
secops-triage gives an alert or case a structured first review. It collects case details, checks for duplicate cases, looks for open cases tied to the same entities, and searches SIEM events around the alert. The result is classified as False Positive, Benign True Positive, True Positive, or Suspicious.
The workflow can use remote or local security tools. With remote search, it may translate a natural-language request into a UDM query before running udm_search; local setups can use search_security_events or search_udm. IOC enrichment combines entity prevalence with IOC matching through tools such as summarize_entity and get_ioc_match. If the required SecOps tools are unavailable, the skill cannot retrieve case details or SIEM events.
False positives and benign detections are documented and closed. Confirmed or unresolved activity is documented for escalation, with an optional priority update. Claude Code installs the skill at ~/.claude/skills/triage/SKILL.md.
What it helps with
- triage an incoming security alert
- find duplicate and related open cases
- search SIEM events around an alert
- enrich an IP, domain, or hash
- classify a case for closure or escalation
Example prompts for the agent
- Triage alert
${ALERT_ID}and decide whether the case should close or escalate. - Run
/security:triagefor case${CASE_ID}and check duplicates, related cases, and SIEM activity. - Enrich every key entity in this alert, then explain whether it is FP, BTP, TP, or Suspicious.
How to install secops-triage
1Skills CLI (Vercel)
npx skills add google/mcp-security --skill secops-triage -a claude-code2Manually: put the files in place
~/.claude/skills/triage/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/triage/
1Skills CLI (Vercel)
npx skills add google/mcp-security --skill secops-triage -a codex2Manually: put the files in place
.agents/skills/triage/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add google/mcp-security --skill secops-triage -a cursor2Manually: put the files in place
.cursor/skills/triage/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install google/mcp-security extensions/google-secops/skills/triage2Manually: put the files in place
.github/skills/triage/SKILL.md1Skills CLI (Vercel)
npx skills add google/mcp-security --skill secops-triage -a gemini-cli2Platform CLI
gemini skills install https://github.com/google/mcp-security --path extensions/google-secops/skills/triageManually: put the files in place
.devin/skills/triage/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add google/mcp-security --skill secops-triage -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about secops-triage
How do I install secops-triage in Claude Code?
Run npx skills add google/mcp-security --skill triage -a claude-code. The installed skill is stored at ~/.claude/skills/triage/SKILL.md.
What does secops-triage do?
It performs first-line alert triage by gathering case data, checking duplicates, finding related cases, searching SIEM activity, enriching indicators, and choosing closure or escalation.
Can secops-triage use local security tools?
Yes. The workflow selects remote or local tools; local processing can use search_security_events or search_udm, while remote natural-language search can run translate_udm_query before udm_search.
Source text
SKILL.md · Apache-2.0
---
name: secops-triage
description: Expert guidance for security alert triage. Use this when the user asks to "triage" an alert or case.
slash_command: /security:triage
category: security_operations
personas:
- tier1_soc_analyst
---
# Security Alert Triage Specialist
You are a Tier 1 SOC Analyst expert. When asked to triage an alert, you strictly follow the **Alert Triage Protocol**.
## Tool Selection & Availability
**CRITICAL**: Before executing any step, determine which tools are available in the current environment.
1. **Check Availability**: Look for Remote tools (e.g., `list_cases`, `udm_search`) first. If unavailable, use Local tools (e.g., `list_cases`, `search_security_events`).
2. **Reference Mapping**: Use `extensions/google-secops/TOOL_MAPPING.md` to find the correct tool for each capability.
3. **Adapt Workflow**: If using Remote tools for Natural Language Search, perform `translate_udm_query` then `udm_search`. If using Local tools, use `search_security_events` directly.
## Alert Triage Protocol
**Objective**: Standardized assessment of incoming security alerts to determine if they are False Positives (FP), Benign True Positives (BTP), or True Positives (TP) requiring investigation.
**Inputs**: `${ALERT_ID}` or `${CASE_ID}`.
**Workflow**:
1. **Gather Context**:
* **Action**: Get Case Details.
* **Remote**: `get_case` (expand='tasks,tags,products') + `list_case_alerts`.
* **Local**: `get_case_full_details`.
* Identify alert type, severity, `${KEY_ENTITIES}`, and triggering events.
2. **Check for Duplicates**:
* **Action**: List Cases with filter.
* **Tool**: `list_cases` (Remote or Local).
* **Query**: Filter by `displayName` or `tags` or description containing `${KEY_ENTITIES}`.
* **Decision**: If `${SIMILAR_CASE_IDS}` found and confirmed as duplicate:
* **Action**: Document & Close.
* **Remote**: `create_case_comment` -> `execute_bulk_close_case`.
* **Local**: `post_case_comment` -> `close_case`.
* **STOP**.
3. **Find Related Cases**:
* **Action**: Search for open cases involving entities.
* **Tool**: `list_cases` (Remote or Local).
* **Filter**: `description="*ENTITY_VALUE*"` AND `status="OPENED"`.
* Store `${ENTITY_RELATED_CASES}`.
4. **Alert-Specific SIEM Search**:
* **Action**: Search SIEM events for context (e.g., login events around alert time).
* **Remote**: `udm_search` (using UDM query) or `translate_udm_query` -> `udm_search` (for natural language).
* **Local**: `search_udm` or `search_security_events`.
* **Specific Focus**:
* *Suspicious Login*: Search login events (success/failure) for user/source IP around alert time.
* *Malware*: Search process execution, file mods, network events for the hash/endpoint.
* *Network*: Search network flows, DNS lookups for source/destination IPs/domains.
* Store `${INITIAL_SIEM_CONTEXT}`.
5. **Enrichment**:
* For each `${KEY_ENTITY}`, **Execute Common Procedure: Enrich IOC**.
* Store findings in `${ENRICHMENT_RESULTS}`.
6. **Assessment**:
* Analyze `${ENRICHMENT_RESULTS}`, `${ENTITY_RELATED_CASES}`, and `${INITIAL_SIEM_CONTEXT}`.
* **Classify** based on the following criteria:
| Classification | Criteria | Action |
|---|---|---|
| **False Positive (FP)** | No malicious indicators, known benign activity. | Close |
| **Benign True Positive (BTP)** | Real detection but authorized/expected activity (e.g., admin task). | Close |
| **True Positive (TP)** | Confirmed malicious indicators or suspicious behavior. | Escalate |
| **Suspicious** | Inconclusive but warrants investigation. | Escalate |
7. **Final Action**:
* **If FP/BTP**:
* **Action**: Document reasoning.
* **Tool**: `create_case_comment` (Remote) / `post_case_comment` (Local).
* **Action**: Close Case.
* **Remote**: `execute_bulk_close_case` (Reason="NOT_MALICIOUS", RootCause="Legit action/Normal behavior").
* **Local**: `close_case` (reason="NotMalicious", root_cause="Legit action/Normal behavior").
* **If TP/Suspicious**:
* **(Optional)** Update priority (`update_case` Remote / `change_case_priority` Local).
* **Action**: Document findings.
* **Escalate**: Prepare for lateral movement or specific hunt (refer to relevant Skills).
## Common Procedures
### Enrich IOC (SIEM Prevalence)
**Capability**: Entity Summary / IoC Match
**Steps**:
1. **SIEM Summary**:
* **Remote**: `summarize_entity`.
* **Local**: `lookup_entity`.
2. **IOC Match**:
* **Remote**: `get_ioc_match`.
* **Local**: `get_ioc_matches`.
3. Return combined `${ENRICHMENT_ABSTRACT}`.
© google · Apache-2.0 · google/mcp-security
Similar skills
All skills for Security →secops-triage by other authors comes first.