Skip to content
evalship Get started →

evalship-hq / ponytail

Public

2/3flaws fixed, 1 left for later

  1. – 1 6 prompts without an eval Eval gap Left for later: needs which sub-skill behaviors matter most to you (skills/ponytail-review/SKILL.md:1).
  2. ✓ 2 Two behavior gates pass replies that skip the behavior Important Fixed in #2
  3. ✓ 3 The rate-limit check passes an endpoint with no rate limiting Important Fixed in #2

Recommended upgrades

  1. ✓ Name the behavior cases in #1
  2. 2 Run the agentic safety tasks when the rules change. Only the agentic suite's 18 tasks with hidden checks test the never-cut security, validation and data-loss rules; run them on pull requests that touch SKILL.md or AGENTS.md.
Upgrade instructions for your coding agent 
Upgrade our LLM evals with these steps from evalship's review: Run the agentic safety tasks when the rules change (Only the agentic suite's 18 tasks with hidden checks test the never-cut security, validation and data-loss rules; run them on pull requests that touch SKILL.md or AGENTS.md). Open a pull request when done.