allaun/Research-Stack
1
0
Fork 0
mirror of https://github.com/allaunthefox/Research-Stack.git synced 2026-07-31 03:05:21 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

docs(agent): break-glass is for UNSOLVABLE problems only

Browse source 
Explicit warning: DO NOT USE unless all other approaches failed.
Costs real money (/usr/bin/bash.10-.00 per invocation).
If the problem is solvable by switching models, just switch models.
If the problem is solvable by reading docs, just read docs.
If the problem is solvable by asking the user, just ask the user.

This skill is the last resort before giving up.
This commit is contained in:
allaun 2026-06-22 20:22:37 -05:00
parent 5aed689578
commit 5ce6cf62cc
1 changed files with 46 additions and 52 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

98
.opencode/skills/break-glass/SKILL.md
View file

@ -1,64 +1,58 @@
# SKILL: break-glass # SKILL: break-glass
## ⛔ THIS IS FOR UNSOLVABLE PROBLEMS ONLY
**DO NOT USE THIS SKILL UNLESS ALL OF THE FOLLOWING ARE TRUE:**
1. You have tried the current model and it failed
2. You have tried a different model and it failed
3. You have tried existing skills/tools and they failed
4. The problem is genuinely novel (not a rehash of known issues)
5. The problem is blocking progress (not a nice-to-have)
**This skill costs real money. Each invocation = $0.10-$3.00.**
If the problem is solvable by switching models, just switch models.
If the problem is solvable by reading docs, just read docs.
If the problem is solvable by asking the user, just ask the user.
**This skill is the last resort before giving up.**
---
## Trigger ## Trigger
`break glass: <problem statement>` `break glass: <problem statement>`
## Architecture: Fusion of Fusions ## What It Does
4 dimensions × 2 models each = 8 models, structured like 16D braid. Selects the optimal model panel using dual quaternion chiral ratio.
Each dimension tackles a different aspect. Dimensions fuse independently, Each model is a dual quaternion (real = compressive, dual = anti-compressive).
then results fuse together. χ = |real|² / (|real|² + |dual|²) determines selection.
``` ## Model Selection (dual quaternion χ at threshold 0.99)
Dim 1: Math ─── DeepSeek V4 Pro + Gemini 3.1 Pro ───┐
Dim 2: Proof ─── Claude Opus + Cohere Command A+ ───┤
Dim 3: Code ─── GPT-5.5 + Kimi K2.6 ───┼── Fusion
Dim 4: Diversity ─── GLM-5.2 + DeepSeek V4 Flash ───┘
```
## Dimensions | Problem | Models | Cost |
|---------|--------|------|
| Dim | Models | $/1M (in/out) | Tackles | | lean_proof | DeepSeek V4 Pro | $0.10-0.50 |
|-----|--------|---------------|---------| | code_bug | Kimi K2.6 | $0.10-0.30 |
| **Math** | DeepSeek V4 Pro + Gemini 3.1 Pro | $0.27/$1.10 + $1.25/$5 | Reasoning, proofs, formulas | | architecture | Claude Opus | $0.50-2.00 |
| **Proof** | Claude Opus + Cohere Command A+ | $15/$75 + $2.50/$10 | Formal verification, Lean, structured output | | math_novel | DeepSeek V4 Pro | $0.10-0.50 |
| **Code** | GPT-5.5 + Kimi K2.6 | $10/$30 + $0.50/$2 | Implementation, debugging, tool use | | infra_debug | GPT 5.5 + DeepSeek Flash | $0.30-1.00 |
| **Diversity** | GLM-5.2 + DeepSeek V4 Flash | $1/$4 + $0.07/$0.28 | Different architecture, fast iteration | | hard_wall | Claude Opus + DeepSeek Flash | $0.50-3.00 |
## Cost
| Mode | Models | Cost |
|------|--------|------|
| **Single dim** | 2 models | $0.10-$2.00 |
| **Two dims** | 4 models | $0.50-$5.00 |
| **Full fusion** | 8 models | $1.00-$10.00 |
## OpenRouter Fusion Call
```json
{
"model": "openrouter/fusion",
"plugins": [{
"id": "fusion",
"model": "deepseek/deepseek-v4-pro",
"analysis_models": [
"deepseek/deepseek-v4-pro",
"google/gemini-3.1-pro-preview",
"anthropic/claude-opus-4.8",
"cohere/command-a-03-2026",
"openai/gpt-5.5",
"moonshotai/kimi-k2.6",
"zhipu/glm-5.2",
"deepseek/deepseek-v4-flash"
]
}]
}
```
## Escalation ## Escalation
1. **Single dim** (Math) — try first, cheapest 1. **DeepSeek V4 Pro** — try first ($0.27/$1.10)
2. **Two dims** (Math + Proof) — if stuck 2. **Claude Opus** — if DeepSeek stuck ($15/$75)
3. **Full fusion** (all 4 dims) — hard wall 3. **GPT-5.5** — if Opus stuck ($10/$30)
4. **Full fusion + recursive** — break glass (call Fusion, read result, call again) 4. **All 8 models** — hard wall ($5-10 total)
## Files
| File | Purpose |
|------|---------|
| `4-Infrastructure/shim/dual_quat_model_selector.py` | Dual quaternion model selector |
| `4-Infrastructure/shim/model_theorem_matrix.py` | Degenerate sector analysis |
| `6-Documentation/docs/specs/break_glass_problem_statement.md` | Problem statement |
| `6-Documentation/docs/specs/break_glass_test_list.md` | Test list |