ennheng/hermes-thariq-workflow — explained in plain English
Analysis updated 2026-05-18
Run a structured blind-spot scan before starting a complex feature.
Have the agent interview you to clarify a vague task before it starts building.
Generate several different implementation approaches before committing to one.
Track deviations from a plan while an agent implements a feature.
| ennheng/hermes-thariq-workflow | 0xkinno/neuralvault | 0xlocker/d17-contracts | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | — | TypeScript | Solidity |
| Setup difficulty | easy | hard | hard |
| Complexity | 2/5 | 4/5 | 5/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires an existing Hermes agent setup, installed by cloning into the Hermes skills directory.
This project is a skill for the Hermes AI agent framework that encodes a structured 7-phase process for working with AI on coding tasks. The idea behind it is that once an AI model is capable enough, the main obstacle to good results is no longer the AI itself, but how well a person can identify and communicate what they don't yet know about the problem they're solving. The workflow has seven named phases, each triggered by saying a specific phrase to the agent. The Blind Spot Pass phase asks the agent to find gaps in your thinking. The Brainstorm phase generates four different approaches to a problem. The Interview phase flips the usual pattern: instead of you asking the agent questions, the agent asks you one question at a time to clarify what you want. The References phase anchors the work in real code or design examples. The Plan phase produces an implementation plan that leads with the decisions most likely to change later. The Implement phase builds the solution while tracking any deviations from the plan. The final Deliver and Quiz phase packages the result and then tests your own understanding of it. The README frames the workflow using four categories: known knowns, known unknowns, unknown knowns, and unknown unknowns, and describes the whole process as converting the unknown categories into known ones. To use it, you clone the repository into a specific folder inside your Hermes skills directory, then say a trigger phrase in a Hermes session to activate it. The README recommends using this for complex multi-file features, unfamiliar codebases, new technology areas, or debugging failed attempts with unclear causes, and recommends against using it for single-line bug fixes or fully scoped, trivial tasks. The project is released under the MIT license.
A Hermes AI agent skill that walks you through a structured 7-phase workflow for tackling complex coding tasks.
Use freely for any purpose, including commercial use, as long as you keep the copyright notice.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.