methodical-hermes-skill-tool-builder

What problem does it solve?

It solves the problem of Hermes creating skills and Tool SOPs that are inconsistent, unsafe, or hard to validate—causing brittle routing, unclear action boundaries, and verification gaps that lead to real runtime failures.

Core Features & Use Cases

• Agent-native skill construction: Converts repeated agent behaviors, workflows, and procedures into reusable runtime skills with clear invocation contracts.
• Tool SOP hardening for procedures: Guides Hermes to produce tool- and SOP-shaped artifacts with guardrails for credentials, privacy, destructive actions, and approval thresholds.
• Progressive disclosure design: Ensures SKILL.md stays lean while deeper details live in support files so Hermes loads only what it needs.
• Evals and verification discipline: Requires stop rules, expected outputs, and verification/eval coverage proportional to risk so “done” is evidence-backed.
• Artifact classification and boundary routing: Helps Hermes decide whether a request should become an Agent Skill, Tool SOP, Review Protocol, human playbook/SOP, or a lightweight draft.

Quick Start

Use methodical-hermes-skill-tool-builder to turn your repeated tool workflow into a Hermes Tool SOP with explicit stop/ask rules and verification steps.