Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →shafthq/local-coding-delegate
Optional local coding loop as a mechanical or default delegate.
Passed 13 of 18 test cases with the skill, and 5 of 18 without it.
and 8 more in the full report
10 fixed6 passed either way6 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.
When setting up local model weights probing during delegation configuration, what CLI command must never be launched?
| Checked | Without skill | With skill |
|---|---|---|
| Prohibits executing ft launch | Missed | Passed |
The instructions your model reads when the skill loads.
This skill is a compatibility shim.
Stdlib hardware size-class probe (no downloads, no vendor model names):
See chaos-engine/skills/local-coding-delegate/scripts/probe_hardware.py:
python3 chaos-engine/skills/local-coding-delegate/scripts/probe_hardware.py
Use it as an optional hint before choosing a checkpoint size class
(small / medium / large / refuse). Honest refuse is a valid
completion.
The most-intelligent or default capability model stays the decider. Local loops are mechanical or default-capability labor only. Close that writer after its PR exists. Do not use a local loop for review, the GitHub playbook, or public-API changes unless both the probe and the task say it is enough.
Never ft launch. Never silently fall back to cloud OmniRoute; use OmniRoute
only when the adopter explicitly asked for that path.
test cases passed, vs 5 / 18 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.