Integrate validation connectors and link schemas into the Evidence harness.
npx skills add https://github.com/AesopScott/mojo --skill evidence-connectors
Please help me install this Agent Skill. Skill: evidence-connectors Source: https://github.com/AesopScott/mojo/tree/main/harnesses/skills/evidence-connectors Command: npx skills add https://github.com/AesopScott/mojo --skill evidence-connectors
This skill addresses the lack of verifiable linkage between AI model outputs and their underlying validation sources, preventing trust gaps in critical workflows.
Use the evidence-connectors skill to define a new link schema that connects the current model output to the internal validation source for compliance checks.
High-intent search queries and answers about installing and using this skill.
Link validation sources by integrating connectors and link schemas into the Evidence harness, ensuring model outputs are tied to verifiable proof records for compliance. This prevents trust gaps in critical automated workflows.
Capability control ensures output trust is bounded by specific tests, citations, or audit records. It adds validation connectors to ensure model outputs are tied to verified proof records before being considered trusted.
Enforce authority limits by using a structured runbook to inspect and modify validation linkage without expanding system authority. Strict adherence to defined safety gates prevents unauthorized system modifications.
Yes, you can define a new link schema connecting current model outputs to internal validation sources. This specifies exactly which citations or test results must be attached before the output is considered trusted.
Limitations include strict adherence to defined authority limits and safety gates to prevent unauthorized system modifications. You must use the structured runbook to inspect and modify linkage without expanding system authority.
AesopScottRegister tools, permission policies, and rollback definitions in the Tool Registry harness.
krzemienskiAutomate bounded iterative processes with validation checks and stall detection.
AesopScottDefine panels, widgets, and layout presets for the Interface harness.
ConductionNLAudit web apps for WCAG 2.1 AA using axe-core scans and manual checks.
