gensecaihq/wazuh-autopilot/ioc-enrichment

IOC Enrichment

Turn raw indicators into decisions: block, watch, or ignore. NIST CSF 2.0 ID.RA (threat intelligence informs risk) and DE.AE (context for adverse events).

Procedure

1. get_case — list indicators with role attacker or observed that are public:

public IPs, internet domains, URLs, file hashes.

1. For each (max 20 per run, attacker-role first):

• check_ioc_reputation(indicator=...) — reputation verdict and sources.
• analyze_security_threat(indicator=..., indicator_type=...) — threat context.
• search_security_events(query=<indicator>, srcip=<ip>, time_range="7d") — internal

sightings (how many hosts?). For hashes, check whether the Wazuh VirusTotal integration already scored the file: rule 87105 (level 12) means positives were found; rule 87104 means none; rules 87101, 87102 are integration errors (rate limit, credentials), not verdicts. get_wazuh_alerts(rule_groups=["virustotal"]) lists them. Load wazuh-malware-detection for detail.

1. Optionally search_external_context for public reporting — see egress rules.
2. Grade each source with the Admiralty Code; combine into a verdict.

Egress rules for search_external_context

search_external_context sends the query outside the environment (third-party web search). Only send:

• Public IPs, public domains/URLs, file hashes, CVE ids, malware/actor names.

Never send: internal hostnames, internal IPs (RFC 1918 etc.), usernames, email addresses, case ids, customer names, file paths, or raw log lines. If unsure, don't send. Respect TLP: indicators received as TLP:RED or TLP:AMBER+STRICT must not be queried externally.

Admiralty Code (NATO / STANAG 2511)

Source reliability: A completely reliable · B usually reliable · C fairly reliable · D not usually reliable · E unreliable · F cannot be judged. Information credibility: 1 confirmed by other sources · 2 probably true · 3 possibly true · 4 doubtful · 5 improbable · 6 cannot be judged.

Typical grading: commercial/curated feed with history → B; single open-source blocklist → C; one-off paste or forum post → D/F. Internal Wazuh sightings of confirmed attacks → A1.

Pyramid of Pain (value of blocking)

Hash values (trivial for attacker to change) < IP addresses < Domain names < Network/host artifacts < Tools < TTPs (hardest to change). Prefer recommending detections on artifacts/TTPs; treat IP blocks as short-lived (24h–7d).

Verdict per indicator

Shared infrastructure warning: cloud/CDN IPs (large providers) host many tenants — blocking them can cause outages; state this explicitly.

Output

• add_entities updating each indicator's enrichment: verdict, sources with Admiralty

grade, first/last seen internally, internal host count, TLP.

• add_finding titled IOC enrichment with the per-indicator table and recommendation.

standard_refs: NIST-CSF-2:ID.RA, NIST-CSF-2:DE.AE.

• update_case confidence if intelligence materially changes it.