Corvus Benchmarks

What it does

When your agent uses it

• : running benchmarks
• Interpreting results
• Regenerating benchmark models
• Troubleshooting BDN issues

Example prompts

• “/corvus-benchmarks”

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

1. 1Always clean Job-* directories before running. Stale ones cause file locks; BDN silently drops benchmarks.
2. 2Never pipe BDN output through truncating commands (Select-Object -First N). This kills the host process.
3. 3filter '*' must be single-quoted in PowerShell to prevent glob expansion.
4. 4buildTimeout 1200 is required — source generators make default 120s too short.
5. 5Detect completion by polling for result files, not by waiting on shell output

What it can do on your machine

Read from SKILL.md and the folder at commit b9e3040. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • dotnet
  • node

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Corvus Benchmarks loads about 1.2k tokens when it runs. Until then it costs about 146 tokens; SKILL.md has 302 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~146

When it runs · the whole SKILL.md, loaded when a task matches

~1.2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from corvus-dotnet/Corvus.JsonSchema at commit b9e3040, republished under its Apache-2.0 licence (© corvus-dotnet). 302 words, ~1,240 tokens.

name

corvus-benchmarks

description

Run, interpret, and maintain BenchmarkDotNet benchmarks for JSON Schema validation and query languages. Covers the B/ (frozen baseline) vs C/ (current) directory convention, stale Job-* cleanup, --buildTimeout, result file polling, regenerating C/ models after codegen changes, and JSONata/JMESPath/JsonLogic/JSONPath benchmarks. USE FOR: running benchmarks, interpreting results, regenerating benchmark models, troubleshooting BDN issues, adding new benchmark schemas. DO NOT USE FOR: general .NET performance analysis (use the analyzing-dotnet-performance skill).

# 1. Build in Release
dotnet build Corvus.Text.Json.slnx -c Release -v q

# 2. Run correctness tests first
dotnet test --solution Corvus.Text.Json.slnx -c Release -f net10.0 --filter "TestCategory!=failing&TestCategory!=outerloop"

# 3. CRITICAL — Clean stale Job-* directories
$benchDir = "benchmarks\Corvus.Text.Json.Benchmarks"
Remove-Item "$benchDir\bin\Release\net10.0\Job-*" -Recurse -Force -ErrorAction SilentlyContinue
Remove-Item "$benchDir\BenchmarkDotNet.Artifacts\results\*" -Force -ErrorAction SilentlyContinue

# 4. Run benchmarks
cd $benchDir
dotnet run -c Release -f net10.0 -- --filter '*' --buildTimeout 1200

# Clean C/ first
Remove-Item -Recurse -Force benchmarks\Corvus.Text.Json.<Name>BenchmarkModels\C\*

# Regenerate
dotnet run --project src\Corvus.Json.CodeGenerator -f net10.0 -c Release -- `
    <schema-path> `
    --rootNamespace Corvus.<Name>Benchmark.Current `
    --outputRootTypeName <Name>Schema `
    --outputPath benchmarks\Corvus.Text.Json.<Name>BenchmarkModels\C `
    --engine V5

© corvus-dotnet, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .github/skills/corvus-benchmarks of corvus-dotnet/Corvus.JsonSchema.

Open the folder on GitHubat commit b9e3040