Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →majiayu000/policyengine-github-agent-skill
Guidance for working with the PolicyEngine GitHub agent bot
Passed 14 of 21 test cases with the skill, and 6 of 21 without it.
and 8 more in the full report
10 fixed5 passed either way6 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.
Draft the exact error response payload generated when an external contributor mentions the PolicyEngine agent on an issue.
| Checked | Without skill | With skill |
|---|---|---|
| The response states that only PolicyEngine/core-developers team members may invoke the bot. | Missed | Passed |
The instructions your model reads when the skill loads.
The PolicyEngine GitHub agent is an automated bot that can be invoked on GitHub issues and pull requests using @policyengine mentions. It helps with code reviews, bug fixes, and implementing features across PolicyEngine repositories.
NEVER use '@policyengine' or '@policyengine-auto' in bot responses. These mentions trigger the bot and create infinite loops where the bot repeatedly responds to itself.
When the bot posts a comment containing @policyengine, GitHub notifies the bot account, which triggers another invocation. This creates a chain reaction of bot responses that can generate dozens or hundreds of comments before being stopped. See PR #22 for a real example of this issue.
Instead of mentioning the bot directly:
❌ Don't do this:
✅ Do this instead:
The bot can only be invoked by members of the PolicyEngine/core-developers team. Non-members will receive a permission error.
When the bot responds to issues or PRs:
test cases passed, vs 6 / 21 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.