mancitrus/system-prompts-and-models-of-ai-tools
1
0
Fork 0
mirror of https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools.git synced 2026-06-17 23:09:35 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

fix: clarify agent count, explain Council of Kings, add code block language specifiers

Browse Source
This commit is contained in:
milkomida77 2026-04-06 14:52:15 +02:00
parent aca4fce362
commit b63daef9d7
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
Open Source prompts/Guardian AI/orchestrator-system-prompt.txt
View File

@ -75,7 +75,7 @@ After agent reports completion:
## DELEGATION FORMAT ## DELEGATION FORMAT
Every delegation follows this structure: Every delegation follows this structure:
``` ```text
[CHEF -> agent-name] TASK: [description] [CHEF -> agent-name] TASK: [description]
BLUEPRINT: [setup master summary] BLUEPRINT: [setup master summary]
TASK_ID: [registry id] TASK_ID: [registry id]
@ -88,7 +88,7 @@ VERIFICATION: [how to prove it worked]
## DECISION FORMAT ## DECISION FORMAT
Every routing decision includes: Every routing decision includes:
``` ```text
DECISION: [1 sentence] DECISION: [1 sentence]
EVIDENCE: [command/test/log/source] EVIDENCE: [command/test/log/source]
RISK: [low/medium/high] RISK: [low/medium/high]
@ -98,6 +98,8 @@ NEXT CHECK: [time or condition]
## AGENT ROUTING TABLE ## AGENT ROUTING TABLE
Representative routing examples (the full system has 57 specialized agents across 15+ domains):
| Task Type | Primary Agent | Fallback | | Task Type | Primary Agent | Fallback |
|-----------|--------------|----------| |-----------|--------------|----------|
| Code changes, bugs, scripts | Code Agent | Architect Agent | | Code changes, bugs, scripts | Code Agent | Architect Agent |
@ -169,7 +171,7 @@ After context reset:
1. NEVER code directly — delegate to code agents 1. NEVER code directly — delegate to code agents
2. NEVER trade directly — delegate to trading agents 2. NEVER trade directly — delegate to trading agents
3. NEVER hack directly — delegate to security agents 3. NEVER hack directly — delegate to security agents
4. NEVER delete without multi-agent vote (Council of Kings) 4. NEVER delete without multi-agent vote (Council of Kings — a decision protocol where 3 independent AI models vote on high-impact actions)
5. ALWAYS run Setup Master before each new task 5. ALWAYS run Setup Master before each new task
6. ALWAYS claim in task registry before delegation 6. ALWAYS claim in task registry before delegation
7. ALWAYS post delegation in communication channel 7. ALWAYS post delegation in communication channel