Verifying & Explaining Evidence
When HardKAS operates, it leaves a trail of evidence in the .hardkas/artifacts/ Canonical Store. The CLI provides tools to interrogate this evidence interactively.
1. Verify Workspace Integrity
hardkas verify
What it does:
- Scans all JSON artifacts in
.hardkas/artifacts/. - Validates each payload against its official schema.
- Recomputes the
contentHashand ensures it matches the payload perfectly. - Ensures the
parentArtifactIdandrootArtifactIdlinks are unbroken (i.e., no missing causal dependencies).
Contract: verify only checks cryptographic and schema integrity of the files on disk. It does not validate transactions against a live Kaspa node. When verifying a single artifact, it will attempt a deterministic replay audit if supported by the environment, but in recursive workspace mode it strictly verifies DAG continuity and semantic integrity.
2. Explain an Artifact
hardkas explain <artifactId | path>
What it does: Produces a human-readable, deterministic explanation of a specific event. HardKAS reads the artifact, identifies its schema, traverses its lineage backward, and prints out the causal story.
Resolution Contract:
You must provide either the exact 64-hex artifactId (the Canonical Identity) or the workspace-relative path to the .json file. You cannot use a txId or contentHash here, due to strict namespace separation.
3. Why did this happen? (Lineage Graph)
hardkas why <artifactId | path>
What it does:
Similar to explain, but focuses purely on the causal graph. It walks the parentArtifactId references all the way to the root of the workflow (the initial Intent/Plan) and prints the unbroken chain of events that resulted in this artifact.