# `@dharma-ai-labs/agent-fabric-evidence-reduction`

Policy-bound local reduction for provider trajectories. It creates bounded,
redacted capsules plus deterministic coverage, completion, runtime, and
tool-discipline signals while keeping complete evidence in the encrypted local
vault by default.

These signals select sessions for deeper review. They cannot determine factual
grounding, inference validity, sycophancy, or a semantic remediation without
approved content evidence.

- [Evidence boundary](https://github.com/dharma-ai-labs/dharma-agent-fabric#cognitive-integrity-evidence-ladder)
- [Security and privacy](https://github.com/dharma-ai-labs/dharma-agent-fabric/blob/main/docs/10-security-privacy-and-threat-model.md)
- [Source](https://github.com/dharma-ai-labs/dharma-agent-fabric/tree/main/packages/evidence-reduction)
- [Dharma AI](https://www.dharma-ai.io)

Licensed under MIT.