Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
9 results for “agent skill”
SkillEffect: Checked Lowering for Memory-Bounded Agent Tools
SkillEffect is a new checked-lowering runtime system that enforces memory bounds during AI agent tool execution by verifying program transformations before execution, enabling reliable operation under fixed memory caps.
Aug 19, 2026
What Is a Skill Worth? Structure-Aware Shapley Valuation of Agent Skills
Researchers introduced SkillSV, a structure-aware Shapley valuation framework to assign credit to internal components (e.g., rules, scripts, heuristics) of AI agent skills—enabling more faithful, actionable, and explainable skill analysis under fixed agents and task distributions.
Aug 6, 2026
Ponytail Agent Skill Corrects Its Own Benchmark After Contributor Challenge
Ponytail, a single-author GitHub repository of instruction files (not executable code), revised its headline claim of 80–94% code reduction after community challenge revealed benchmark flaws—replacing it with a lower, agentic-run figure of 54%.
Aug 5, 2026
Self-Supervised Skill Optimization
Researchers introduced Self-Supervised Skill Optimization (SSO), a method that improves LLM agent skills using only unlabeled task data and an LLM judge—no ground-truth labels, rewards, or external evaluators—demonstrating competitive performance against supervised methods.
Aug 3, 2026
Agent Skill to Force Docs in ASD-STE100 Simplified Technical English
A Hacker News thread discusses an experimental agent skill that enforces documentation compliance with ASD-STE100 Simplified Technical English, a controlled language standard used in aerospace and defense technical writing.
Jul 31, 2026
PatchOptic for Shared-State LLM Workflows with Projected Views and Verified Structured Updates
PatchOptic is a new interface for LLM agentic workflows that enforces structured, verified updates to shared state using projected views and patch contracts — addressing the gap between local model edits and global state consistency.
Jul 9, 2026
SkillCloak Lets Malicious AI Agent Skills Evade Static Scanners with Self-Extracting Packing
Researchers demonstrated that static scanners designed to detect malicious 'skills' for AI coding agents can be reliably evaded using self-extracting packing techniques, and proposed a runtime-based detection alternative.
Published Jul 6, 2026 · Analyzed Jul 8, 2026
Show HN: QUALITY.md – open format/specification, agent skill, and CLI
A community-submitted open specification for agent skill documentation called QUALITY.md was posted to Hacker News, introducing a lightweight format for describing AI agent capabilities and interfaces.
Published Jul 2, 2026 · Analyzed Jul 5, 2026
Cloudflare Ships Agent Skills for Zero Trust Deployment and Migration
Cloudflare released an open-source library of agent skills to automate Zero Trust deployment and migration, including interoperability with Zscaler and Palo Alto Networks, leveraging logic from its proprietary Descaler program.
Published Jun 25, 2026 · Analyzed Jul 4, 2026