SKILLEMALL.ai
GitHub ↗ RU Sign in

CC hugging-face-evaluation

Add and manage evaluation results in Hugging Face model cards. Supports extracting eval tables from README content, importing scores from Artificial Analysis API, and running custom model evaluations with vLLM/lighteval. Works with the model-index metadata format.

synthetic-sciences/OpenScience Hermes author: synthetic-sciences Apache-2.0 14 files · 8 scripts body ≈ 5 580 tokens Open the sourcegithub.com↗ analyzed 5 d ago

Add and manage evaluation results in Hugging Face model cards.

As a process C 53/100 · Has gaps — References files that are not bundled: ...

AnalyzerAI and agentstype and topics are labelled automatically from the skill text
Runs in: Hermes Agent
JSON
Technical rating
C
82/100
safety, quality, tests
Safety 60%
98
Quality 40%
59
Run on models
none yet
Process rating
C
53/100
Has gaps
References files that are not bundled: ...
Tools and files w 18
0
Result and completion w 14
0
Progress reporting w 2
0
the three weakest of ten parameters · all ten

How to improve

  1. For Hermes the description must be one sentence under 60 characters; move the conditions to a "When to Use" section.
  2. The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
  3. 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 · 2

✓ No critical or high findings

Medium and low: 2
  • low Exfiltration read-dotenv examples/USAGE_EXAMPLES.md:24
    Reads a .env file (test fixture / example file)
    cp examples/.env.example .env
    fixture
  • low Exfiltration net-credential-use examples/USAGE_EXAMPLES.md:347
    Credential used in a network call (verify the destination is the intended service) (test fixture / example file)
    curl -H "x-api-key: $AA_API_KEY" \
    fixture

Files scanned: 13. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • warning description-long-hermes description is 264 chars; the Hermes authoring standard requires ≤ 60 (one sentence, ending with a period)
  • warning body-long SKILL.md body ≈ 5580 tokens (recommended < 5000); move details to references/
  • warning missing-ref reference to a missing file: ...
  • note frontmatter-key unknown frontmatter key "dependencies"

Process rating: all ten parameters 53/100

Will not run. References files that are not bundled: ...
  • 0Tools and files. 1 referenced file(s) missing: ...
  • 0Result and completion. Does not say what the result is
  • 0Progress reporting. Says nothing while it works
  • 50When it triggers. No condition that starts the skill
  • 70Inputs and preconditions. Inputs and preconditions are listed
  • 70Execution cost. Instruction body is 5580 tokens
  • 85Steps. 121 steps, 1 vague phrases
  • 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
  • low 10 top-level sections: this looks like several domains in one skill

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

  • +5Description has no quoted example phrases that should trigger the skill
  • +4Description does not say when NOT to use the skill (false activations)
  • +3Output format is not stated: the model decides each time
  • -31 of 7 scripts are never mentioned in SKILL.md
  • +2Single-language instructions
  • +3Description length 264: enough signal without eating the budget
  • +4Structure: 36 headings
  • +3Step-by-step instructions: 121 items
  • +4Has examples (29 code blocks)
  • +1License stated

Quality base 70; lint remarks subtract, signals add up to 100. Result: 59.