Coming soon
The postpeg Agent Skill
The postpeg Agent Skill teaches a coding agent every endpoint, each network’s rules, errors and retries, so the code it writes against the API works the first time.
What it teaches your agent
An Agent Skill is a folder of instructions a coding agent such as Claude Code loads when a task needs it. The postpeg skill covers:
- Every endpoint, with the exact request and response fields
- Each network’s rules: text limits, media, required options
- Publishing flow: check, publish or schedule, then follow each account’s status
- Errors by
code, what to retry, rate limits andIdempotency-Key - When to ask before acting: publishing, replying and sending are public
Coming soon
The skill is written and lives in the postpeg repository; there’s no public install link yet. Until then, point your agent at llms-full.txt (the whole docs in one file), or connect the MCP server, which works today.
Questions
- How do I install it?
- There’s no public install link yet. The skill is written and will be published with an install command; this page will have it the day it ships.
- Skill or MCP?
- Use the skill when a coding agent writes code that calls the API; use MCP when an assistant should act for you. The MCP server works today.
Other integrations
Send your first post in a few minutes
The 7-day trial has every feature and 3 connected accounts, with no card. The quickstart takes you from a key to a published post.