Skip to content
Get packs

Test Reasoning Serialization

Checks that reasoning settings are sent correctly to each AI provider's API

Source Download skill
Repository
tailcallhq/forgecode
Stars
7,638
Licence
Apache-2.0
Files
2 files

What it does

When installed, the agent can run a test suite that builds the forge tool in debug mode, sends test requests to several AI providers (OpenRouter, Anthropic, GitHub Copilot, Codex), captures the outgoing request body, and checks that reasoning/thinking settings appear in the expected JSON fields. It also confirms that invalid effort values are rejected before any request is sent. It is meant for maintainers verifying the forge codebase rather than for general end users.

Great for

  • Contributors to the forge repository changing how reasoning config is serialized
  • Maintainers checking provider-specific JSON after an API change
  • Developers debugging why a reasoning setting is not reaching a provider

Skip it if

  • General users who only want to use forge; the skill tests internals and is not needed to run the tool
  • Running against production accounts casually: real provider requests may be made and could incur cost, and the script reads credentials from the environment
  • Anyone expecting it to test other providers or features beyond reasoning fields

Try asking

  • “Run the reasoning serialization tests for this repo”
  • “Verify that ReasoningConfig effort maps to reasoning.effort for OpenRouter”
  • “Check the Anthropic thinking budget_tokens field after my change”
  • “Run a single manual reasoning test for codex gpt-5.1-codex”
Step-by-step guide: how to use Test Reasoning SerializationInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the Apache-2.0 licence and an attribution note.

    test-reasoning.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip test-reasoning.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip test-reasoning.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run the script from the repository root, since it uses relative paths like target/debug/forge
  • Read scripts/test-reasoning.sh before running it, and use a test account or low-cost keys
  • Use the single-test command first to inspect the captured request before running the full suite

What's inside

  • scripts/test-reasoning.sh
  • SKILL.md

Redistributed unmodified from tailcallhq/forgecode under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.