Open sourceJavaScript · ≈180k stars · MIT
Ponytail
A minimalism skill: your agent thinks like the laziest senior dev in the room.
Ponytail is one prompt, a SKILL.md with a compact AGENTS.md version, that pushes coding agents to write less code. It installs as a Claude Code plugin or a Codex plugin with two lifecycle hooks, and any agent that reads a rules file can use it.
Pros & cons
Each point cites its sources and says what kind of evidence backs it. A second reviewer tried to refute every claim before it was kept.
Pros
- DocsThe author's benchmark (39 tasks, Opus 5.5, 5 runs each) reports 53% less code, 45% fewer tokens and 26% lower cost, with all 30 runs of the six security tasks passing.[1]
- DocsIts stated rule is to never cut validation, error handling, security or accessibility, and it reports 98% of risky logic shipping with a test, against 68% without it.[1]
- DocsVersion 5.3 added a Claude Code band that shows cost, tokens and per-turn spend.[1]
- User reportsOne commenter finds its basic simplicity heuristics useful for getting agents to make quick, small edits.[1]
Cons
- DocsAll headline numbers come from the author's own benchmark rather than independent replication.[1]
- User reportsSkeptics say it is a handful of rules plus a lot of plugin boilerplate, replaceable by asking 'Is there a simpler solution?'[1][2]
- User reportsUsers asked for measurements on standard benchmarks (SWE-Bench Pro, Terminal Bench) and objected to it adding 'ponytail:' markers to code comments and commit messages.[1][2]
- User reportsHarness drift: the OpenCode plugin used a V1-only API and did nothing on OpenCode 2 (September 2026, closed).[1]
- Best for
- Teams fighting agent over-engineering who want smaller diffs without dropping safety checks.
- Watch out for
- Check the claimed savings on your own repo before relying on them.
Assessed Oct 29, 2026 · adversarially reviewedSuggest a correction
Neighbours on the workbench
MDX · ≈307k stars · MIT
obra/superpowers
An agentic skills framework and software methodology that works.
Maintained by Jesse Vincent (obra)
Shell · ≈306k stars · MIT
mattpocock/skills
Small, composable agent skills for real engineering work.
Maintained by Matt Pocock (AI Hero)
JavaScript · ≈292k stars · MIT
affaan-m/ECC
Agent harness performance optimization system: skills, instincts, memory and security for coding agents.
Maintained by Affaan Mustafa
Connections
Relations from the Harness Hub knowledge graph, as charted in the State of Harness reports.
Maintained by · 1
- DietrichGebertOrganisation
Built on · 2
- Claude CodeProject
- CodexProject
Similar entries · 1
- Andrej Karpathy SkillsProject
Sources
Verified Oct 29, 2026
