Blocked
The defence prevented or interrupted the activity before the objective was met.
Morgana · Validation platform
Turn validation objectives into evidence.
Morgana plans and runs authorised validation, gathers execution and detection evidence, and helps teams understand what worked, what was missed and what needs to improve.
Free to download
Morgana is free to download from Camelot, the X3M.AI community hub on GitHub. Get the software and documentation without requesting access.
How it works
From an objective to a reasoned conclusion
An analyst moves from an objective, to a plan, to test results and a reasoned conclusion — with evidence at every step.
- 01Scope the behaviour
Choose relevant Families and Variants — the exact behaviours worth validating.
- 02Plan the strategy
Pick a deliberate execution strategy and freeze it into a reproducible plan.
- 03Run controlled tests
Execute authorised tests through enrolled agents and gather output and telemetry.
- 04Interpret the evidence
Reason over detection evidence: timing, behavioural match, telemetry quality and uncertainty.
Detection assurance
Separate findings, not a single score
A blocked action, a successful execution and a detected behaviour are different findings. Morgana keeps them distinct so assurance reflects what the evidence actually shows.
Executed
The activity completed as planned — the outcome to correlate against your detections.
Detected
Telemetry shows the behaviour was observed and identified as it happened.
Repeatability & learning
Evidence that improves the next test
Reviewed outcomes inform improvements and subsequent tests. Each cycle produces evidence that feeds the next, turning validation into a sustained practice rather than a one-off exercise.
Explore Continuous ValidationFree to download
Get Morgana without requesting access
Morgana is free to download from Camelot, the X3M.AI community hub on GitHub. Run the Windows installer, follow the guide and get started — no contact or approval required to obtain the software.
The installer is Morgana-Server-Setup.exe in theInstall folder. Morgana must still be used only for authorised operations on systems covered by explicit written approval.
Questions or collaboration? You can reach the founder onLinkedIn.
Morgana's cognitive core
Intelligence built into the validation loop
Morgana joins an executive coordination layer, specialist cognitive software agents and a local inference engine to support planning and evidence reasoning — always with human authority over what runs.
Morgana Brain
The Executive Brain coordinates objectives, context and specialist reasoning — the software layer that connects what should be validated with what the evidence shows.
Multi-agent reasoning
Specialist cognitive agents cooperate on defined analytical tasks through the Cognitive Router — distinct from the endpoint agents that execute tests.
Micro-modeling
Small, specialised models handle focused tasks with task-specific evaluation, using the smallest model that reaches the required quality.
Morgana Intelligence Engine (MIE)
The local inference path runs models on your own host. Running the validated local configuration means no external AI egress; external providers require explicit configuration.
What AI actually helps with
- Planning a validation objective — turning scope into a concrete plan.
- Assessing evidence quality — interpreting telemetry and reasoning over what it shows.
- Distinguishing prevention from detection — a blocked action and a detected behaviour are separate findings.
- Identifying what to investigate or retest — surfacing gaps and priorities from evidence.