13 KiB
Colab Template 📓
Problem Map 3.0 Troubleshooting Atlas
Reusable template for Colab-based fix contributions
Quick links:
- Back to Templates Hub
- Back to Contribution Checklist
- Back to Fix Recipe Template
- Back to Prompt Template
- Back to Community Fix Lab
- Back to Official Fixes
- Back to Atlas landing page
- Back to Atlas Hub
- Get the Atlas Router TXT
This file is the reusable template for writing a clean Colab or notebook-based fix contribution.
If the Fix Recipe Template helps you document a repair asset broadly, this template helps you document runnable or inspectable notebook-based repair assets in a form that is easier to reproduce, compare, and review ✨
Use this template when you want to submit:
- one Colab demo
- one repair notebook
- one comparison notebook
- one replay notebook
- one small benchmark rerun notebook
The goal is not to dump an .ipynb file.
The goal is to keep notebook contributions:
- routed
- runnable
- inspectable
- scoped
- honest about limits
Quick start 🚀
Use this template in the following order:
- fill Title
- fill 0. Quick summary
- fill 1. Notebook type
- fill 2. Atlas routing context
- fill 3. Problem this notebook addresses
- fill 4. Intended use
- fill 6. Notebook sections
- fill 7. Baseline failure and 8. First repair move
- fill 9. Expected outputs and 11. Misrepair warning
- fill 13. Files included, 14. Runtime and dependency notes, and 15. Limitations
- add 12. Optional WFGY escalation only if it is genuinely relevant
If you want the fastest possible start, jump to:
- 17. Copy-paste mini skeleton
Short version:
one notebook
one broken baseline
one first repair move
one before / after comparison 📌
Required vs optional quick map 🗂️
| Section | Status | Purpose |
|---|---|---|
| Title | Required | identify the notebook clearly |
| 0. Quick summary | Required | explain what the notebook does |
| 1. Notebook type | Required | classify the notebook |
| 2. Atlas routing context | Required | anchor the notebook in atlas logic |
| 3. Problem this notebook addresses | Required | define the target case |
| 4. Intended use | Required | show when and why to use it |
| 5. Required inputs | Required | show what the run needs |
| 6. Notebook sections | Required | define the notebook structure |
| 7. Baseline failure | Required | show the broken starting point |
| 8. First repair move | Required | define the intervention |
| 9. Expected outputs | Required | define what success looks like |
| 10. Suggested evaluation fields | Optional | add metrics only if useful |
| 11. Misrepair warning | Required | prevent wrong notebook use |
| 12. Optional WFGY escalation | Optional | only if deeper bridge is relevant |
| 13. Files included | Required | make the submission reviewable |
| 14. Runtime and dependency notes | Required | make the notebook runnable |
| 15. Limitations | Recommended | keep the contribution honest |
| 16. One-line maintainer note | Required | help quick review |
A good notebook contribution usually looks like this 🌱
A strong notebook contribution usually has:
- one family
- one clear case
- one broken baseline
- one first repair move
- one before / after comparison
- one expected success signal
- one honest limit statement
Small and legible beats huge and messy almost every time.
Important notebook discipline 🔒
A notebook contribution should not try to do everything at once.
If one notebook is trying to be:
- a giant benchmark
- a full production replica
- a universal proof
- a routing engine
- a deep WFGY research pack
- and a polished demo
all at once, it is usually too broad.
Choose the main notebook role first.
Make the before / after logic visible.
Keep the run easy to inspect.
Title
Replace this line with a short clear title.
Example:
F4 Readiness Gate Demo - Minimal closure repair notebook
0. Quick summary
Write 1 to 3 short sentences.
Example:
This notebook demonstrates a workflow failure caused by missing readiness checks.
It compares the broken baseline with a repaired version that adds a simple readiness gate.
1. Notebook type
Choose one or more, but choose the main notebook role first.
- demo notebook
- benchmark rerun
- repair notebook
- trace notebook
- reproduction notebook
- comparison notebook
If your notebook could fit more than one label, choose the main role first and use the others only if they genuinely describe the same notebook.
2. Atlas routing context ⭐
This is one of the most important sections in the whole template.
A notebook without routing context is much harder to review and much easier to misuse.
Primary family
F?
Secondary family
F? or None
Broken invariant
Write one short sentence.
Best current fit
Write the nearest node, family entry, or edge-fit wording.
Why this notebook belongs here
Write 2 to 4 short sentences.
3. Problem this notebook addresses
Describe the specific problem this notebook demonstrates.
Useful questions:
- what is broken in the baseline
- why a notebook is useful here
- what first repair move is being tested
- what this notebook is not trying to solve
Keep this short and concrete.
4. Intended use ⭐
State clearly how this notebook should be used.
Examples:
- first public demo
- contributor reproduction
- benchmark sanity check
- teaching notebook
- route-first repair demonstration
Optional format:
Use stage
...
Target user
...
Expected runtime
...
This section matters because a good notebook is not only about running code.
It is also about when and why someone should open it.
5. Required inputs
List the minimum inputs.
Examples:
- corpus
- query set
- schema file
- workflow config
- notebook parameters
- baseline JSON
- expected JSON
Use a short format like:
Input A:
Input B:
Input C:
6. Notebook sections ⭐
A good notebook should usually contain these sections, even if some are very small.
Section A · Setup
What the notebook loads or installs.
Section B · Case framing
What case is being shown and how it routes in the atlas.
Section C · Broken baseline
Show the failure first.
Section D · First repair move
Apply the family-level first repair move.
Section E · Re-run
Run the repaired version.
Section F · Comparison
Show before / after differences.
Section G · Optional WFGY escalation
Show how the same case could be explored more deeply if needed.
7. Baseline failure ⭐
Describe the broken baseline.
Useful questions:
- what is the baseline workflow
- what fails first
- what visible symptom appears
- why is this failure interesting
Optional mini format:
Baseline setup
...
Broken behavior
...
Failure note
...
A notebook contribution is usually much easier to understand when the broken starting point is visible.
8. First repair move ⭐
Describe the first repair move being tested.
Useful questions:
- what intervention is being applied
- why it is the correct first move
- what should improve after the intervention
Optional mini format:
Repair action
...
Why this is first
...
Expected improvement
...
9. Expected outputs ⭐
Describe what the notebook should produce.
Examples:
- before / after metrics
- clearer trace output
- better schema pass rate
- fewer wrong anchors
- successful closure
- improved support rate
Optional format:
Before
...
After
...
Success signal
...
A good notebook output is not only a metric.
It should also make the before / after difference legible.
10. Suggested evaluation fields
List only useful fields.
Examples:
support_rateclosure_successschema_pass_ratetrace_completenesswrong_anchor_ratefield_loss_count
Optional mini block:
Metric 1:
Metric 2:
Metric 3:
Only include fields that are actually useful.
11. Misrepair warning ⭐
This section is required.
Wrong first move
...
Why it is tempting
...
Why this notebook is not meant for that
...
This helps prevent notebooks from teaching the wrong reflex.
12. Optional WFGY escalation
Use this only if deeper WFGY exploration is relevant.
When to escalate
...
What should be passed into WFGY
- routed family
- broken invariant
- first repair already attempted
- unresolved pressure
What WFGY is expected to add
...
Do not use this section to skip atlas routing.
13. Files included
List the files included.
Example:
demo.ipynbinput.jsonexpected_output.jsonREADME.md
14. Runtime and dependency notes ⭐
Be explicit.
Examples:
- Python version
- notebook runtime type
- external packages
- API keys if needed
- GPU not required
- internet required or not required
Optional mini format:
Colab link
...
Replay-only or live
...
Runtime type
...
Dependencies
...
API key required
Yes / No
Short, clear notes are enough.
15. Limitations
Be honest.
Examples:
- toy dataset only
- small case only
- one model family only
- partial demo, not full system benchmark
- notebook proves first move, not full closure
16. One-line maintainer note
Write one short line that helps review the contribution.
Example:
Small F4 closure demo notebook with before / after run and simple success metric.
17. Copy-paste mini skeleton ✂️
Use this when you want the fastest possible start.
# Title
## 0. Quick summary
...
## 1. Notebook type
...
## 2. Atlas routing context
Primary family:
Secondary family:
Broken invariant:
Best current fit:
Why this notebook belongs here:
## 3. Problem this notebook addresses
...
## 4. Intended use
Use stage:
Target user:
Expected runtime:
## 5. Required inputs
...
## 6. Notebook sections
A.
B.
C.
D.
E.
F.
G.
## 7. Baseline failure
...
## 8. First repair move
...
## 9. Expected outputs
...
## 10. Suggested evaluation fields
...
## 11. Misrepair warning
Wrong first move:
Why it is tempting:
Why this notebook is not meant for that:
## 12. Optional WFGY escalation
...
## 13. Files included
...
## 14. Runtime and dependency notes
...
## 15. Limitations
...
## 16. One-line maintainer note
...
Next steps ✨
After this page, most readers continue with:
- Back to Contribution Checklist
- Back to Templates Hub
- Back to Fix Recipe Template
- Back to Prompt Template
If you want to return to the broader product surface:
If this template helps your workflow, consider:
- starring the WFGY repo
- opening an issue
- shipping one small clean notebook contribution 🌟
18. Closing note
A good notebook contribution does not need to be huge.
It only needs to be:
- routed
- runnable
- scoped
- inspectable
- honest about limits