cloudcast
/cloudcast
Pomaga rozwiązywać i rozwijać benchmark routingu transmisji w wielu chmurach Cloudcast.
- Autor
- openJiuwen-ai ↗
- Repozytorium
- FAST-Evolve ↗
- Licencja
- Apache-2.0
- Ostatnia zmiana repozytorium
- 13.11.2026
Ocena i bezpieczeństwo
- Pliki w paczce
- 7
- Skrypty
- 6
- Komentarze HTML
- 0
- Adresy URL w SKILL.md
- 0
Brak sygnałów. Repozytorium zmienione 13.11.2026, bez odwołań do wycofanych modeli i API.
Apache-2.0, otwarta: można używać i zmieniać, z zachowaniem licencji i informacji o zmianach.
- 6 skryptów w paczce. Przejrzyj je przed instalacją, bo Claude może je uruchomić.
Skrypty: initial_program.py, scripts/broadcast.py, scripts/download_data.py, scripts/evaluator.py, scripts/simulator.py, scripts/utils.py
Co robi i kiedy
Pomaga rozwiązywać i rozwijać benchmark routingu transmisji w wielu chmurach Cloudcast.
Oryginalny opis (EN)
Use when the user wants to solve or evolve the Cloudcast multi-cloud broadcast routing benchmark.
Jak działa
- Evaluator
- Program interface
- Optimization objective and guidance
- Required data
Podgląd SKILL.md
---
name: cloudcast
description: Use when the user wants to solve or evolve the Cloudcast multi-cloud broadcast routing benchmark.
---
# Cloudcast Multi-Cloud Broadcast Optimization
Evolve a routing algorithm that broadcasts data from one cloud region to several
destinations at minimum total transfer cost. The evaluator runs the candidate on
five network configurations and maximizes the normalized fitness
`1 / (1 + total_cost)`.
## Evaluator
Use the standard attempt-local artifact contract:
```text
main_artifact_name = solution.py
artifact_ref = runs/<run_id>/tasks/<task_id>/attempts/<attempt_id>/solution.py
```
In the commands below, `<skill_dir>` is the directory containing this SKILL.md.
Run only the benchmark evaluator command:
```bash
python <skill_dir>/scripts/evaluator.py --run-id <run_id> --task-id <task_id> --artifact-ref <artifact_ref> --out runs/<run_id>/tasks/<task_id>/attempts/<attempt_id>/eval_result.json
```
The evaluator writes the normalized result with `status`, `artifact_ref`,
`fitness`, `metrics`, `profile`, and `misc` fields. A successful evaluation has
exit code 0 and `status: "success"`.
## Program interface
The candidate must define:
```python
def search_algorithm(src, dsts, G, num_partitions):
...
```
It must return a `BroadCastTopology`-compatible object whose paths contain every
destination and every partition, with contiguous graph edges carrying the
original edge data. `initial_program.py` is the initial reference program.
## Optimization objective and guidance
You are an expert in cloud infrastructure optimization. Your task is to evolve
the `search_algorithm(src, dsts, G, num_partitions)` function to minimize overall
data transfer cost across multiple clouds.
Focus on efficiently broadcasting input data to multiple destination nodes by
leveraging parallel paths and overlapping transfers across networks. Use the
`BroadCastTopology` class and `make_nx_graph` function to identify low-cost
routes. Prioritize strategies that reduce redundant transfers, balance load
across networks, and exploit multi-network topologies to minimize cost.
## Required data
The following source-controlled-by-download files are required under this skill:
- `profiles/cost.csv`
- `profiles/throughput.csv`
- `examples/config/intra_aws.json`
- `examples/config/intra_azure.json`
- `examples/config/intra_gcp.json`
- `examples/config/inter_agz.json`
- `examples/config/inter_gaz2.json`
Download them once with:
```bash
python <skill_dir>/scripts/download_data.py
```
The evaluator uses only the benchmark data and the candidate artifact.
Pliki w paczce (7)
- SKILL.md 3 KB
- initial_program.py 4 KB
- scripts/broadcast.py 1 KB
- scripts/download_data.py 1 KB
- scripts/evaluator.py 17 KB
- scripts/simulator.py 8 KB
- scripts/utils.py 3 KB
Wersje
W bazie nie ma innych skilli o nazwie cloudcast.
Migawka bazy z 13.11.2026 · wersja paczki 9c4ee1792966 (skrót SHA-256 całej paczki)
Pobrania w ostatnich 30 dniach: 0 · ranking
Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.