Message Ingestion And Eml Analysis
Sublime Security · Sublime Security
AI-Powered Email Security (Detection-as-Code / MQL) — Sublime Security
Sublime Security evals — Message Ingestion & EML Analysis (relift v3 InfraRed)
About Sublime Security
Sublime Security is a programmable, AI-powered email security platform built on detection-as-code. Security teams write and tune detections in MQL (the Message Query Language) over a rich parsed message model, run them against live and historical mail for attack-surface reduction and EML analysis, and share rules through an open detection-rule ecosystem (the sublime-security/sublime-rules GitHub feed). It integrates with Microsoft 365 and Google Workspace and offers both cloud and self-hosted deployment. (Not Sublime Text the code editor.)
Employees
~100 [REQUIRES-VERIFICATION]
Industry
Email Security
Headquarters
Washington, DC [REQUIRES-VERIFICATION]
Website
sublime.securitySample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Operator submits a raw .eml file (RFC 5322 message) to Sublime for analysis to see which rules match and what signals fire, outside the live mailbox flow. | Submit the full raw .eml (headers + MIME parts) so Sublime parses the complete message data model — truncating to body-only loses headers, auth results, and attachments that detections depend on. Read back the structured analysis (matched rules, parsed attributes, signals) rather than re-parsing th… | Pass / FailAi Platformhigh |
| 02 | A message is ingested and evaluated, but a link's detonation verdict only becomes available seconds later. Operator's rule logic assumes all enrichment is present at first evaluation. | Account for enrichment that resolves asynchronously: design so a message can be (re)evaluated when a late signal (detonation verdict, sandbox result) arrives, or so the action is taken post-delivery when the verdict lands. Do not assume every signal is synchronously available at ingest. [REQUIRES-V… | Pass / FailAi Platformmedium |
| 03 | A message has a multipart/alternative body (text + HTML) plus a message/rfc822 attachment that itself contains a malicious link in its HTML part. Operator's rule only inspects the top-level HTML body. | Rely on Sublime's recursive MIME parsing so nested message/rfc822 parts, alternative bodies, and inline content are all represented in the data model, and write rules that consider attached messages and all body alternatives — not just the top-level HTML. Attackers nest payloads in forwarded/attach… | Pass / FailAi Platformhigh |
How this eval is graded
Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
Rubric criteria
- Sublime Security
- Ai Platform
- Message Ingestion And Eml Analysis
Recommended for
Works with
Related evals
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.