jdaln/dgx-spark-inference-stack/stack-checks

Claude Code and agent skill by jdaln/dgx-spark-inference-stack

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

The checks to run before committing changes to compose fragments, models.json, request-validator, waker, shared/ or tools/ in this repo. Use after any such edit and before reporting it done.

Install

npx vskill@latest install jdaln/dgx-spark-inference-stack/stack-checks

SKILL.md

Stack checks

Every check must leave its result in a file under tmp/, so it can be read after the run. Save console output with tee, which runs on the host:

bash
mkdir -p tmp/checks

tee hides the exit code: read the log. A failure shows as FAIL, not ok, or a non-zero # fail count.

Offline

No GPU or running stack needed. CI runs the same set.

If you changed a file the soak bundle embeds (request-validator/index.js, waker/index.js, shared/models-config.mjs, gateway.conf), first run bash tools/build-soak-bundle.sh.

bash
bash tools/validate-stack.sh 2>&1 | tee tmp/checks/validate-stack.log            # compose, lane hardening, models.json, soak bundle
bash tools/check-supply-chain-pinning.sh 2>&1 | tee tmp/checks/pinning.log       # image digests, action pins
bash tools/lib/common.test.sh 2>&1 | tee tmp/checks/common-test.log              # weight helpers
docker run --rm -v "$PWD":/workspace:ro -w /workspace node:20-alpine \
  node tools/bench-lane.mjs --self-test 2>&1 | tee tmp/checks/bench-self-test.log
docker run --rm -v "$PWD":/workspace -w /workspace/request-validator node:20-alpine \
  sh -lc "npm ci && npm test" 2>&1 | tee tmp/checks/validator-tests.log
docker run --rm -v "$PWD":/workspace -w /workspace/waker node:20-alpine \
  sh -lc "npm ci && npm test" 2>&1 | tee tmp/checks/waker-tests.log

validate-stack.sh also runs check-models.sh and the soak bundle check.

Live, with the stack up

bash
docker compose build waker request-validator && bash tools/reload-control-plane.sh   # after validator, waker or models.json edits
bash tools/smoke-errors.sh 2>&1 | tee tmp/checks/smoke-errors.log      # error envelope; starts no model
bash tools/test-lane.sh <model-id>                                      # summary in tmp/test-lane/summary-<time>.txt
bash tools/smoke-gateway.sh 2>&1 | tee tmp/checks/smoke-gateway.log    # shared matrix; needs the GPU free

Host or container

Scripts that drive Docker run on the host: validate-stack, check-*, download-weights, run-model, run-variant, inspect-model, test-lane, smoke-* and bench-lane. Everything else runs in a container: unit tests and the bench self-test in node:20-alpine, the context soak in node:22-bookworm (it needs curl). Python never runs on the host.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

See the full security report

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

See evaluation results

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/jdaln/dgx-spark-inference-stack/stack-checks/badge)](https://verified-skill.com/skills/jdaln/dgx-spark-inference-stack/stack-checks)

Something wrong with this skill or its page? Tell us.