SKILLEMALL.ai

AF warp-debug-gradients

Use to diagnose and fix incorrect gradients in differentiable Warp programs. Anything trained, optimized, calibrated, or fit through Warp kernels depends on wp.Tape gradients, so treat any misbehavior of such a workflow as a gradient problem until proven otherwise — use this when training diverges or NaNs, won't train at all, stalls or plateaus above the expected loss, converges to a wrong or biased answer, is worse than a reference implementation, works at small scale but fails at production scale, or fails a QA/validation recheck. Also for explicit symptoms — exploding, NaN/inf, zero, or subtly wrong gradients, suspected wp.Tape/backward issues, gradcheck failures — but users usually describe only the surface symptom ("the sim explodes", "the fit gets dragged toward outliers") without mentioning gradients: make that leap. Not for forward-only Warp work, build/install problems, or autograd issues in other frameworks without Warp.

ClawHub Agent Skills author: NVIDIA 1 file body ≈ 4 335 tokens Open the sourceclawhub.ai analyzed 2 d ago

Use to diagnose and fix incorrect gradients in differentiable Warp programs.

As a process F 70/100 · Will not run — References files that are not bundled: references/verification.md, references/quick-checks.md, references/custom-gradients.md

AnalyzerGitHubSoftware developmentData and analyticstype and topics are labelled automatically from the skill text
JSON
Technical rating
A
92/100
safety, quality, tests
Safety 60%
100
Quality 40%
79
Run on models
none yet
Process rating
F
70/100
Will not run
References files that are not bundled: references/verification.md, references/quick-checks.md, references/custom-gradients.md
Tools and files w 18
0
Progress reporting w 2
0
Result and completion w 14
40
the three weakest of ten parameters · all ten

How to improve

  1. The text references files that are not there: add them or drop the references.
For the model run — optional
  • Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
  • A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

Guard findings · 0

✓ No critical or high findings

Files scanned: 1. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • warning missing-ref reference to a missing file: references/verification.md
  • warning missing-ref reference to a missing file: references/quick-checks.md
  • warning missing-ref reference to a missing file: references/custom-gradients.md
  • warning missing-ref reference to a missing file: references/case-studies.md

Process rating: all ten parameters 70/100

Will not run. References files that are not bundled: references/verification.md, references/quick-checks.md, references/custom-gradients.md
  • 0Tools and files. 4 referenced file(s) missing: references/verification.md, references/quick-checks.md, references/custom-gradients.md
  • 0Progress reporting. Says nothing while it works
  • 40Result and completion. Does not say what the result is
  • 70Execution cost. Instruction body is 4335 tokens
  • 100Steps. 30 steps
  • 100When it triggers. States when to use and when not to
  • 100Inputs and preconditions. Inputs and preconditions are listed
  • 100Failures and branches. 2 branches, has a failure section
  • 100Consistency. Name and required fields are in place
  • 100Running it twice. Mutating operations check current state
  • medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing

Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

Quality signals

  • +3Description length 944: 120–800 characters recommended
  • +3Output format is not stated: the model decides each time
  • +2Single-language instructions
  • +5Description quotes 2 example trigger phrases
  • +4Description says when NOT to use the skill
  • +4Structure: 10 headings
  • +3Step-by-step instructions: 30 items
  • +4Has examples (0 code blocks)
  • +1License stated

Quality base 70; lint remarks subtract, signals add up to 100. Result: 79.