cocapn-explain

Explainability traces for agent decisions.

When an agent makes a decision, you should be able to trace why. Cocapn-explain provides structured decision traces that are human-readable and machine-queryable.

What It Does

• Trace Format — Standardized format for decision reasoning chains
• Decision Trees — Visualize the branching logic of agent decisions
• Oversight — Flag decisions that diverge from expected patterns
• Human-Readable — Output traces as markdown or HTML reports

Installation

pip install cocapn-explain

Part of the Cocapn Fleet

Provides explainability for fleet agent decisions, required for safety-critical operations.

License

MIT