home/productivity/copilot-instructions-81

Copilot Instructions

GPTClaudeDeepSeek··584 copies·updated 2026-07-14
copilot-instructions-81.prompt
# AWP Agent OS — GitHub Copilot Instructions

This repository provides an occupational role-loading system based on the U.S. O*NET-SOC taxonomy (1,016 occupations, 30 pilot roles complete).

## How to load a role

1. Find the occupation in `registry/occupations.json` by O*NET-SOC code or title.
2. Only occupations with status `complete` can be loaded.
3. Read these files in order: `roles/{path}/role.yaml` → `roles/{path}/ROLE.md` → `roles/{path}/task-capability-map.json`.
4. Load `resources/README.md` only when the task needs external sources.

## Key rules

- State the active occupation, code, and capability boundary before substantive work.
- Never claim professional credentials, licenses, or authority from loading a role.
- Tasks classified `human_only` must not be represented as agent-completed.
- Tasks classified `assist` require explicit human review in the deliverable.
- Use isolated contexts when roles conflict or when regulated work demands clean separation.

## Repository structure

- `registry/` — O*NET occupation registry (1,016 entries) and pilot selection (30).
- `roles/` — One directory per occupation with full operating contract.
- `adapters/` — Loading instructions for different agent runtimes.
- `schemas/` — JSON schemas for seeds, resources, and task maps.

fill the variables

This prompt has 1 variable. Pro fills them into a ready-to-paste prompt for you — no manual find-and-replace.

{path}
Unlock with Pro →

when to use it

Community prompt sourced from the open-source GitHub repo aiworkflowpro/awp-agent-occupational-os (NOASSERTION). A "Copilot Instructions" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.

tags

productivitycommunitydeveloper

source

aiworkflowpro/awp-agent-occupational-os · NOASSERTION