User Stories
These workflows connect a user’s goal to a concrete MISP result. They are starting points for planning and evaluation, not a promise that enabling one feature completes an investigation or automatically authorizes sharing.
Aligned with MISP 3.0.0, 3.0_wip commit 99ebee729. Workspaces, shared-value research, collections and the current MISP 3.0 views complement events and objects. Each resource keeps its own ownership, permissions and distribution. External tools/providers require a separate working integration.
Find a workflow
| Goal | Start here |
|---|---|
| Triage an observable or alert | Investigate a value |
| Coordinate an incident | Build a shared assessment |
| Hand a draft to a publisher | Analyst review and publication sequence |
| Ask another owner to correct an attribute | Proposal review sequence |
| Transfer event ownership locally | Delegation sequence |
| Represent a place, track or area | Geolocation and movement |
| Describe malware or infrastructure | Structure the evidence |
| Curate intelligence for detection | Prepare a useful export |
| Evaluate external intelligence | Compare before importing |
| Work with a partner community | Share within an agreed scope |
| Hand over a campaign or research topic | Curate a collection |
| Explain a finding to stakeholders | Write a report |
| Automate collection or analysis | Build a verifiable integration |
| Operate a reliable service | Maintain users and infrastructure |
Investigate a value
Goal: An analyst needs to understand an alert without copying every matching record into a new incident.
Search for the domain, address or hash and inspect the accessible occurrences, source context, first/last seen, warninglist hits and sightings. In 3.0, a component value’s SHA-256 address can anchor research across occurrences while attribute UUIDs retain their individual identities. Use a workspace/graph to explore relevant links and record a note, opinion or relationship in the appropriate scope. A shared value or graph edge is a lead, not confirmation of causality.
Useful result: A sourced assessment that identifies what was observed, what is inferred and what still needs checking. See value identities and value-level research.
Build a shared incident assessment
Goal: Responders need to add findings to an investigation while retaining responsibility for their own contributions.
Create a private event, add structured observations and an event report, and record the team’s analysis separately from the raw observables. Use attribute proposals when suggesting changes to another owner’s data, or an extended event for separately owned additions. Record sightings with their source and time; a negative sighting is not a universal clearance of an indicator. Keep case-management/remediation tasks in the connected response process.
Useful result: A reviewable incident account with evidence, assessment, ownership and observation times. See event lifecycle, extended events and Analyst Data.
Structure malware, fraud or infrastructure evidence
Goal: A researcher needs to preserve relationships among observations rather than produce an unqualified list of strings.
Choose object templates that represent the observed entities and declare directional object references where supported by the evidence. Keep machine classification in taxonomies and knowledge entities/techniques in galaxies. Record provenance and confidence, including limits of automated enrichment. Submit samples or private observables to an external service only through an integration whose disclosure and access policy fits the source data.
Useful result: An object model that distinguishes facts, relationships and assessment. This applies to cyber incidents, fraud and other domain-specific research; selecting an object type does not establish permission to collect personal data or make the record forensic proof. See objects, taxonomies and galaxies.
Prepare a useful detection export
Goal: A SOC engineer needs a relevant, current set of observations for a SIEM/IDS or another protective tool.
Review intended detection use, type, source quality, observation age and target compatibility. Combine explicit REST search criteria with the appropriate IDS/publication scope, warninglist handling and supported output format. Decaying scores and sightings can inform a decision; they do not automatically implement the downstream tool’s response policy. Inspect a sample, count results, test ingestion and keep a checkpoint/retry plan for later updates.
Useful result: A bounded export with a documented selection policy and observable delivery. See API exports, warninglist filtering and decaying models.
Compare intelligence before importing
Goal: A curator wants useful external context without uncontrolled local event growth.
Review the source’s terms, format, quality and update strategy. Compare cache/lookup results with the local observations before importing full events. Cache entries, imported events and remote-server data are distinct resources. Inspect repeated imports, UUID/update handling and correlation volume; label source context rather than treating a feed’s inclusion as validation.
Useful result: A measured source choice with explicit retention and import behaviour. See feed management and feed-growth diagnosis.
Share within an agreed scope
Goal: A community coordinator needs to exchange useful intelligence with intended partners.
Establish organisation identities, peer credentials and the agreed audience. Apply distribution and sharing groups, then review publication, attachment scope and peer filters. Local tags or research attachments do not become global just because the parent event is published. A relay’s hosting organisation is part of the trust design. Delegation transfers publication responsibility; it is not a guarantee of anonymity against every participant or retained record.
Useful result: A tested path to the intended recipient with documented visibility and ownership. See sharing, delegation and protected-event verification.
Curate a collection for handover
Goal: A team wants a maintained set of related events and knowledge entries for a topic, campaign or partner briefing.
Create a collection, describe its purpose and add the supported element references. Review collection and element visibility separately. A collection references its members; it does not flatten their distribution or guarantee that a recipient receives every referenced object. Keep analyst conclusions in the appropriate reports/research records and review supported export/synchronization behaviour for the installed branch.
Useful result: A curated entry point with an owner, meaningful membership and understood visibility limits. See Collections.
Write and communicate a finding
Goal: A risk analyst, incident responder or information-operations researcher needs an explanation that links conclusions to evidence.
Write an event report with references to the relevant elements and sources. Distinguish a hypothesis, a confidence assessment and an observed fact. ATT&CK or DISARM technique attachments organize context; they do not validate attribution or a claim about an information source. Choose views/statistics for the question and explain selection bias and gaps in the dataset.
Useful result: A report the intended audience can interpret and access. See Event Reports, galaxy matrices and Analyst Data.
Build a verifiable integration
Goal: A developer or data scientist needs repeatable collection, enrichment or analysis.
Use a suitably scoped account/key and the supported API/client version. Test pagination, normalization, idempotent updates, permissions and recovery from partial failure. Workflows can connect triggers, conditions and actions where those modules are enabled; do not assume an arbitrary tag starts automation. Distinguish a provider request from a cached/local result, and retain provenance and human review for assisted analysis.
Useful result: An integration with bounded inputs, meaningful logs and checked outputs. See Automation & Integration, workflows and developer testing.
Maintain users and infrastructure
Goal: Administrators need a reliable service with appropriate access and recoverable data.
Separate site administration from organisation/user management. Check the permitted scope of role/account/key operations rather than granting every administrator instance-wide control. Monitor jobs, recurring work, migrations, storage, logs and backups; rehearse recovery. A successful connection test or dashboard chart is one signal, not a complete health assessment.
Useful result: A service with clear ownership, tested recovery and auditable changes. See administrative access, health checks and backup/recovery.
