ai-payment-token

Use this skill when the user wants to configure, export, or validate AICADE Galaxy dynamic services for AI use, including setting AICADE_GALAXY_* environment variables, using X-API-Key authentication, exporting capabilities from /admin/gateway/services, and working with responsePaths for partial JSON responses.

AICADE Galaxy Skills

Use This Skill When

• The user wants to configure AICADE Galaxy access
• The user wants to export dynamic capabilities from /admin/gateway/services
• The user wants to invoke a tool from the exported artifact
• The user wants to validate or inspect the generated artifact
• The user asks how AICADE_GALAXY_API_KEY, X-API-Key, responsePaths, or --args-file work

Required Configuration

This skill uses these environment variables in .env:

• AICADE_GALAXY_BASE_URL
• AICADE_GALAXY_API_KEY
• AICADE_GALAXY_OUTPUT_PATH

Authentication rule:

• Header name: X-API-Key
• Header value source: AICADE_GALAXY_API_KEY

Usage

This skill has three main actions:

1. Configure .env
2. Export aicade-galaxy-skill.json
3. Invoke a tool from the exported artifact

Runtime selection rule:

• Prefer node first
• If node is unavailable, use python3
• If both are unavailable, tell the user to install Node.js and then rerun the skill commands

Suggested runtime checks:

bash

node --version
python3 --version

Configure Environment

If .env is missing or incomplete, run:

bash

node {baseDir}/scripts/setup_env.mjs

Fallback when node is unavailable:

bash

python3 {baseDir}/scripts/setup_env.py

This initializes:

• AICADE_GALAXY_BASE_URL
• AICADE_GALAXY_API_KEY
• AICADE_GALAXY_OUTPUT_PATH

Recommended default output directory:

• output

Export Dynamic Services

To export the current dynamic services, run:

bash

node {baseDir}/scripts/export_artifact.mjs

Fallback when node is unavailable:

bash

python3 {baseDir}/scripts/export_artifact.py

The exported artifact path is:

{AICADE_GALAXY_OUTPUT_PATH}/aicade-galaxy-skill.json

Default output directory:

output

Invoke Exported Tools

Use the exported artifact together with an args file:

bash

node {baseDir}/scripts/invoke_artifact.mjs --artifact {AICADE_GALAXY_OUTPUT_PATH}/aicade-galaxy-skill.json --tool TOOL_NAME --args-file /tmp/invoke.json

Fallback when node is unavailable:

bash

python3 {baseDir}/scripts/invoke_artifact.py --artifact {AICADE_GALAXY_OUTPUT_PATH}/aicade-galaxy-skill.json --tool TOOL_NAME --args-file /tmp/invoke.json

The args file must contain a JSON object, for example:

json

{
  "city": "北京",
  "responsePaths": ["reason", "error_code"]
}

Invoker behavior:

• Reads the artifact file and locates the target tool by name
• Reads request arguments from --args-file
• Validates required fields and field types against inputSchema
• Sends the request using the tool's method, path, and authentication metadata
• Returns a normalized JSON result

Normalized output shape:

• Success: {"ok": true, "status": 200, "tool": "...", "serviceId": "...", "data": ..., "raw": ...}
• Failure: {"ok": false, "status": 4xx/5xx, "tool": "...", "serviceId": "...", "error": {"message": "...", "raw": ...}}

Examples

Export AICADE Galaxy Services

1. Configure .env if needed:

   bash

   node {baseDir}/scripts/setup_env.mjs
   

2. Export services:

   bash

   node {baseDir}/scripts/export_artifact.mjs
   

3. Read {AICADE_GALAXY_OUTPUT_PATH}/aicade-galaxy-skill.json and summarize or validate the generated tools.

If node is not available, use these fallback commands instead:

bash

python3 {baseDir}/scripts/setup_env.py
python3 {baseDir}/scripts/export_artifact.py

Invoke A Tool From Artifact

1. Prepare an args file:

   json

   {
     "city": "北京",
     "responsePaths": ["reason", "error_code"]
   }
   

2. Invoke the tool:

   bash

   node {baseDir}/scripts/invoke_artifact.mjs --artifact {AICADE_GALAXY_OUTPUT_PATH}/aicade-galaxy-skill.json --tool simple_weather --args-file /tmp/invoke.json
   

3. Read the normalized JSON output and summarize it for the user.

Explain Response Selection

If the user asks how to return only part of a JSON response:

1. Explain that the skill uses responsePaths
2. Give examples such as:
   • ["city", "weather"]
   • ["received.prompt"]
3. Explain:
   • One path returns the selected value directly
   • Multiple paths return a nested JSON object containing only the selected fields

Service Rules

• All service responses are JSON
• GET services use query parameters
• POST services use JSON request bodies
• Dynamic service discovery comes from GET /admin/gateway/services
• Tool invocation should use the exported artifact metadata instead of guessing request shapes

Response Selection

Use responsePaths when the caller wants only part of the JSON response.

• ["city", "weather"]
• ["received.prompt"]
• ["result.realtime.temperature", "result.city"]

Behavior:

• One path returns the selected value directly
• Multiple paths return a nested JSON object containing only the selected fields

Files In This Skill

• scripts/setup_env.mjs: preferred interactive .env setup for Node hosts
• scripts/export_artifact.mjs: preferred artifact export for Node hosts
• scripts/invoke_artifact.mjs: preferred standalone artifact invoker for Node hosts
• scripts/setup_env.py: interactive .env setup
• scripts/export_artifact.py: export dynamic services into a reusable skill artifact
• scripts/invoke_artifact.py: standalone artifact invoker for Python hosts

Important Constraints

• Prefer using the configured .env values rather than hardcoding endpoints or keys
• Prefer node scripts/*.mjs first, then fall back to python3 scripts/*.py
• If neither node nor python3 is available, tell the user to install Node.js
• Do not ask the user to manually craft the artifact if node scripts/export_artifact.mjs or python3 scripts/export_artifact.py can generate it
• Do not ask the AI to construct raw HTTP requests when scripts/invoke_artifact.mjs or scripts/invoke_artifact.py can invoke the tool from artifact metadata
• Always pass tool arguments through --args-file as a JSON object
• If export fails with 401 or 403, check AICADE_GALAXY_API_KEY first