Table of Contents

🇫🇷 Version française

Example README template

A template for examples/**/README.md. Copy the block below into a new example's README.md and fill each section. Delete this preamble and any optional section that does not apply. Keep it short — the point is to let someone run the example without reading the code.

Required sections: What it does · Prerequisites · Required data · Run it · Expected output · Approx. duration & cost.

NB: example READMEs are written in English (they live outside docs/, so the bilingual parity contract does not apply to them) — this template's French twin exists for reference only.

One or two sentences: what this crew produces and why it is interesting.

What it does

  • Process: Sequential | Hierarchical | Parallel | Consensual | Graph | Autonomous
  • Agents: — brief role list
  • Tools: tool_a, tool_b, …
  • Key features: what this example demonstrates (task dependencies, memory, A2A, …)
  • Runner: orkeon CLI

Prerequisites

  • .NET SDK ≥ 10.0.300 (source) — or the .NET 10 runtime (release binary)
  • An LLM profile (see the profile matrix under examples/appsettings/); this example was validated against <provider>.
  • <Any other prerequisite: a running service, an API key beyond the LLM, …>

Required data

If the crew reads input files, list them so the reader knows what to mount. Delete this section if the example needs no input data.

Virtual path Mount flag Purpose
/data/<file> --mount ./data:/data:ro

Run it

Give the exact, copy-pasteable command for each supported way. The orkeon CLI is the default entry point (orkeon run <config>); the always-required "from source" row plus, for showcase examples, a binary and a container row.

With the orkeon CLI (installed release binary or dotnet tool install):

orkeon run examples/<path>/config.yaml \
  --settings path/to/appsettings.local.json \
  --mount ./out:/output:rw

From a source checkout (no install — runs your local code):

dotnet run --project src/scripting/Orkeon.Scripting.Cli -- run \
  examples/<path>/config.yaml \
  --settings examples/appsettings/appsettings.<provider>.local.json \
  --mount ./out:/output:rw

From the container (showcase examples only — entry point is orkeon):

docker run --rm \
  -v "$PWD/out:/output" \
  -v "$PWD/appsettings.local.json:/app/appsettings.local.json:ro" \
  ghcr.io/orkeon/orkeon-runners \
  run examples/<path>/config.yaml \
  --settings /app/appsettings.local.json \
  --mount /output:/output:rw

Flag reference: Run your first example (adjust the relative path once this file lives in an example folder).

Expected output

Describe what a successful run prints and/or writes. If the crew writes a file to /output, name it and describe its shape (e.g. "a Markdown report with an executive summary, thematic sections, and a bibliography").

Approx. duration & cost

  • Duration: ~ min on <provider/model>
  • Cost: ~ LLM calls; rough token / $ estimate if known (say "local model — no API cost" when applicable)
Edit this page