The self-documenting graph behind Ghost Intelligence. Every capability is automatically discoverable, searchable, teachable, documented, linked and visualised.
Self-review is how the system catches its own mistakes.
For users
The system reviews its own recent behaviour. Runs reflective reviews over recent decisions and outcomes. Open Reflection to see the latest self-review.
For developers
Reflection (id: reflection) is a partial Research capability rendered at /pulse/reflective. It runs reflective reviews over recent decisions and outcomes. It builds on: self-assessment, learning.
For investors
Reflection: Self-review is how the system catches its own mistakes. It is partial at 68% confidence, demonstrating real, defensible capability rather than a slide.
Learning objectives
Explain what Reflection does in one sentence.
Describe why Reflection matters to the system.
Open /pulse/reflective and identify the live values it produces.
Trace how Reflection relies on self-assessment, learning.
Examples
Open Reflection to see the latest self-review.
Known limitations
Surface is partial — functional but still thin.
Confidence is 68% — treat outputs as advisory.
Future improvements
Open Reflection to see the latest self-review.
Raise completion to Production Ready with more live coverage.