AI coding agents
Integrating through Claude, Cursor, GitHub Copilot or another AI coding agent? Give it the agent docs. They add what the API reference leaves implicit: the rules a correct integration follows, copy-ready client code in Python and TypeScript, and a test plan. Every environment serves them without authentication:
| File | Path | Use it for |
|---|---|---|
| Index | /llms.txt |
Agents that discover docs via the llms.txt convention |
| Full docs | /llms-full.txt |
One Markdown file: integration guide and API reference |
| Claude skill | /agents/e-parcel-api.zip |
Claude Code, Claude.ai, Claude Desktop |
Connect your agent
Section titled “Connect your agent”Install the skill in your project (use ~/.claude/skills instead to have it
in every project):
curl -fsSL https://api.e-parcel.fi/agents/e-parcel-api.zip -o e-parcel-api.zipunzip -o e-parcel-api.zip -d .claude/skills && rm e-parcel-api.zipStart a new claude session. It loads the skill whenever you work on the
integration; type /e-parcel-api to load it explicitly. On Windows, use
Expand-Archive e-parcel-api.zip -DestinationPath .claude\skills instead of
unzip.
Download
https://api.e-parcel.fi/agents/e-parcel-api.zip
and upload it as a custom skill: Customize → Skills → “+” (on older
layouts Settings → Capabilities → Skills → Upload skill). Code execution
must be enabled.
Or add https://api.e-parcel.fi/llms-full.txt to a Project’s knowledge
instead.
Point the agent at https://api.e-parcel.fi/llms-full.txt, or add this to the
AGENTS.md (or equivalent rules file) in your repository:
## E-Parcel shipping APIBefore changing E-Parcel integration code, readhttps://api.e-parcel.fi/llms-full.txt and follow its golden rules.Develop and test against https://api-test.e-parcel.fi only.A good first prompt
Section titled “A good first prompt”Integrate our order fulfilment with the E-Parcel API using the e-parcel-api skill. Target the test environment; the API key is in the
EPARCEL_API_KEYenvironment variable.