SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution
Hongyi Liu, Haoyan Yang, Tao Jiang, Bo Tang, Feiyu Xiong, Yuyu Luo, Zhiyu Li
arXiv:2605.18401 · 130▲ · cs.CL, cs.AI
View on arXiv →Premium readers get an interactive explainer for this paper — a figure you can poke at, not just read.
What it is
SkillsVote is a lifecycle framework for managing libraries of Agent Skills (structured folders of instructions, scripts, and applicability conditions used by LLM coding agents). It does three things: performs agentic search over skill folders to pick relevant skills before a task, decomposes execution traces into skill-linked subtasks after a task, and gates library updates so only successful, reusable discoveries edit or create skills.
Why it matters
For anyone building coding agents that accumulate reusable experience, the paper shows that dumping every trajectory or every available skill into context hurts more than it helps. The attribution layer (deciding whether success came from the skill, the agent's own exploration, or the environment) is meant to stop weak or misattributed lessons from polluting a growing skill library, which is a real operational concern once you have thousands of skills.
Practical takeaway
If you maintain a skill or memory store for agents, treat pre-task retrieval and post-task writeback as separate gated steps rather than static top-k embedding lookup, and only commit updates backed by verifier-confirmed success. Watch for this pattern showing up in agent harnesses like Codex and Claude skills tooling.
Key result
On Terminal-Bench 2.0 (avg@5 accuracy) online evolution improved GPT-5.2 by 2.6 points, GPT-5.4 mini by 1.1, and GPT-5.5 by 0.9 over a no-skills baseline; the largest offline transfer gain was 7.8 points for GPT-5.2 from a library built on historical Terminal-Bench Pro trajectories. Gains are modest and uneven (skills sometimes regress on harder splits and on GPT-5.5), and the authors attribute the small averages to a heavy-tailed effect where skills help a few tasks a lot and hurt others.
Subscribe
Get the next issue.
Free. One email a week. Unsubscribe any time: no account, no dark patterns.