Downstream linking

SkillDev tools

Build the current NiceEval candidate and install it into an authorized downstream project for dogfooding.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Downstream linking skill

What this skill tells your AI

The instructions your AI receives, as published by niceeval/niceeval in .agents/skills/downstream-link/SKILL.md and read by ahel’s review.

Read downstream discovery and dogfooding, the target repository's nearest AGENTS or README, and the Repository Tools boundary.

For a linking task, run pnpm dev:link --help for current syntax. Inspecting existing downstream results does not require building or installing a candidate.

Confirm the target repository and its existing NiceEval source before writing. When installing, build the current candidate once, verify its identity, and install only into the named downstream. Do not infer permission for paid runs, full benchmarks, result deletion, commits, or pushes.

Report the target, consumed NiceEval identity, and the smallest public-entry dogfood check that was run. Include the install result when linking was part of the task.

Signals

GitHub stars
152
Forks
4
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
downstream-link
Source
github.com/niceeval/niceeval