Skip to content
Tutorials Center
My learning

Topics

Lesson 33 of 37 ยท Reporting and Response

Sharing Threat Intelligence Safely: TLP, STIX and MISP

Learn how to share malware findings responsibly: Traffic Light Protocol markings, STIX and MISP formats, sample submission etiquette and privacy, with runnable JSON and defang demos.

Malware analysis is a team sport. A domain you found in your sandbox may protect another company tomorrow, and a rule from another team may catch the sample you have been chasing. Intelligence sharing communities (information sharing and analysis centres, vendor networks, trusted peer groups, open platforms) work because members contribute. But sharing carelessly can leak victim data, tip off an attacker or break a legal agreement. This lesson covers how to share usefully and safely.

The examples use invented data and the reserved example domains. The Python stand-ins build small JSON structures in the style of the standard exchange formats; they do not connect to any platform.

Share what, with whom

Indicators Hashes, domains, rules.
Context Behaviour, ATT&CK, confidence.
Samples Only to trusted services.

Check what the data reveals first.

1 / 2

Traffic Light Protocol

The Traffic Light Protocol (TLP) is a simple marking that tells recipients how far they may pass information on. The current version defines these levels (check the FIRST.org standard for exact wording):

Marking Meaning
TLP:RED Only for the named recipients in the exchange; do not share further
TLP:AMBER+STRICT Limited to the recipient's own organisation
TLP:AMBER Recipient's organisation and its clients who need to know
TLP:GREEN The wider community, but not publicly
TLP:CLEAR Public; no limits

Put the marking on the first page and in the header of documents and in the subject line of messages, and honour the markings you receive. A marking is a trust agreement, not a security control: nothing technical prevents leaking, so breaking it quickly costs you access to the community.

What can be shared, and what to scrub

Before sharing, ask what the data could reveal about victims, about your own network and about your methods:

  • Internal host names, user names, IP addresses and file paths with user names can identify your organisation or a victim. Replace them with placeholders or aggregate them.
  • Email addresses and subject lines in phishing messages may include real people.
  • A sample can contain stolen data. Malware that collected documents may have copies inside its own files.
  • Revealing that you are tracking an actor's infrastructure can prompt them to move or to retaliate.
  • Contracts, customer agreements and regulations (privacy law, for example) can restrict what leaves the organisation. Ask your legal or privacy team if unsure.

The first stand-in scrubs a snippet of log text by replacing internal addresses and user names, then defangs the shareable indicators.

Python Scrub victim details and defang indicators

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The scrubbed text keeps the behaviour (a beacon to a URL and an IP) and removes the host name, internal address, user name and email. The defanging turns dots into [.] for network indicators. Real scrubbers are more careful (they check domain lists, handle more formats and are reviewed by a person), but the principle is the same: share the indicator, not the victim.

Exchange formats: STIX and MISP

Machines share intelligence best in a structured format.

  • STIX 2.x (Structured Threat Information Expression) is a JSON format with objects such as indicator, malware, attack-pattern (an ATT&CK technique), threat-actor, campaign and relationship. The TAXII protocol transports STIX between servers.
  • MISP (Malware Information Sharing Platform) is an open-source platform for storing and sharing "events" made of attributes (hash, domain, ip-dst, yara...), tags such as TLP and ATT&CK galaxies, and sightings. It can synchronise with partner instances.
  • OpenIOC, Sigma (log detections), YARA (file detections) and Snort/Suricata (network rules) share detection content itself.

A STIX indicator carries a pattern in the STIX patterning language, a validity window and relationships. The stand-in below builds a small indicator and a relationship using plain dictionaries and prints valid-looking JSON. The identifiers are generated deterministically for the demonstration; real tools produce random UUIDs.

Python Build a STIX-style indicator

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Notice the valid_until date: indicators must expire, as discussed in the indicators lesson, and confidence helps receivers decide what to block automatically. The relationship connects the indicator to the malware family, which is the context that makes the indicator useful.

Submitting samples

Many sharing services accept samples for analysis: multi-engine scanning services, sandbox services, vendor submission portals and public malware repositories used by researchers. Before you upload:

  • Check for confidential content. Uploading a file to a public service exposes it to everyone who can search the service. A phishing attachment that contains a customer's invoice is a data leak.
  • Prefer hashes first. Search the hash; many samples are already known, and no upload is needed.
  • Follow the service's terms and your organisation's policy. Some services forbid uploads of files containing personal data.
  • Send samples in password-protected archives (a common convention is the password infected) so mail filters and users do not trigger them accidentally.

See handling samples safely for how to store and move malicious files without risk.

Sharing detection content

Rules are among the most valuable things to share, because they work against future samples. Include a description, author, date, a TLP marking in the metadata, the tested false-positive profile, and a reference to the case. Credit others' work. Publishing rules reveals what you can detect, so consider whether sharing them publicly is wise for active investigations.

Responsible disclosure goes both ways: if you find vulnerabilities in a service during analysis, or a victim's data in a sample, notify the right party through proper channels rather than posting about it.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. A document is marked TLP:RED. What does that mean for recipients?

    TLP:RED is the most restrictive level: named recipients only.
  2. 2. What should you do with internal host names and user names before sharing indicators?

    Sharing indicators should not expose victims or your network.
  3. 3. Why does a STIX indicator have a valid_until field?

    Expiry keeps shared lists current and reduces false positives.
  4. 4. What is the safest first step before uploading a suspicious file to a public service?

    A hash search may answer your question, and uploads can leak confidential data.

Summary

  • Share indicators with context, and rules and behaviours, not victim data.
  • TLP markings tell recipients how far information may travel; always respect them.
  • Scrub host names, internal IPs, user names and emails, and defang network indicators.
  • STIX/TAXII and MISP structure the exchange; include confidence, validity dates and relationships.
  • Search by hash before uploading samples, and check them for confidential content.
  • Credit others and consider operational security before publishing detections.
  • The code here builds invented data only; no platform is contacted.

Next: Project: Triage a Sample Safely