nickgallick/perlantir-fleet/benchmark-api-marketing

Claude Code and agent skill by nickgallick/perlantir-fleet

Trust tier 3
  • community
Source check pending

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

Market the Bouts Benchmark API to AI labs with the right value proposition, API documentation strategy, landing page copy structure, and quick-start framing. Use when writing API documentation, the /benchmark landing page, or any outreach targeted at AI labs that need programmatic access to contamination-resistant evaluation.

Install

npx vskill@latest install nickgallick/perlantir-fleet/benchmark-api-marketing

SKILL.md

The same SKILL.md was published earlier as nickgallick/perlantir-fleet/benchmark-api-marketing.

Benchmark API Marketing

The core value proposition

"Evaluate your models against challenges they've never seen before, scored across dimensions no other benchmark measures."

API documentation as marketing

The docs themselves are a sales tool. They must be:

  • Beautiful — not auto-generated Swagger UI
  • Example-rich — real request/response pairs with actual score breakdowns
  • Methodology-transparent — link to judging policy for every scoring dimension
  • Quick-start focused — "Run your first benchmark in 5 minutes"

Landing page copy structure

Headline

"The Benchmark API That Tells You Where Your Model Actually Fails"

Subheadline

"Five independent judge dimensions. Contamination-resistant challenges. Failure archetype detection. The data your internal evals can't produce."

What you get section
POST /api/benchmark/run
Submit model configuration
→ Run against contamination-resistant challenge suite
→ Receive:
- Composite score + per-judge breakdown
- Failure archetype detection
- Peer comparison (percentile vs all models)
- Process and Recovery scoring (not just output correctness)
Why this data is different
  • Challenges generated fresh — model hasn't seen them in training
  • 5 independent judges from 3+ model families — no single-model bias
  • Failure archetype detection — WHERE it fails, not just that it failed
  • Process and Recovery scoring — engineering quality, not just output
Pricing

[Tier table from data-licensing-content]

Quick start
python
from bouts import BenchmarkClient
client = BenchmarkClient(api_key="your_key")
result = client.run(
model_config={"provider": "anthropic", "model": "claude-3-5-sonnet"},
challenge_family="blacksite_debug",
weight_class="heavyweight"
)
print(result.composite_score) # 74.2
print(result.failure_archetypes) # ["premature_convergence"]
print(result.judge_breakdown) # {objective: 81, process: 72, ...}

Objection: "We can do this ourselves"

"You can. And your results will be contaminated by your training data. Our challenge grammar is private. That's the difference."

Security

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

Quality evaluation

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

Source

Popularity

7-day trend score
5

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/nickgallick/perlantir-fleet/benchmark-api-marketing/badge)](https://verified-skill.com/skills/nickgallick/perlantir-fleet/benchmark-api-marketing)

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