[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$f2C-cuELwVg9wJEHfDTUZq9DeWnj51tU46KaXo42eePA":3},{"category":4,"tools":8,"otherCategories":77},{"name":5,"slug":6,"count":7},"Agent Testing","agent-testing",2,[9,55],{"id":10,"name":11,"slug":12,"description":13,"url":14,"githubUrl":15,"logoUrl":16,"openSource":17,"categories":18,"repo":21,"articles":40,"createdAt":54},"entity_01kzsa6ykgextv2krhnxvx212d","Dynobox","dynobox","Cross-harness testing for multi-step agent flows","https:\u002F\u002Fdynobox.xyz","https:\u002F\u002Fgithub.com\u002Fdynobox\u002Fdynobox","https:\u002F\u002Fr2.minima.ltd\u002Forg_01jakhtww5fk6b0jmxj0qfp9rf\u002Ffavicon-6008b356.svg",true,[19],{"name":5,"slug":6,"count":20},0,{"fullName":22,"url":15,"stars":23,"forks":20,"openIssues":20,"language":24,"license":25,"topics":26,"pushedAt":37,"archived":38,"licenseSpdx":25,"licenseOsi":17,"licenseVerifiedAt":39},"dynobox\u002Fdynobox",10,"TypeScript","Apache-2.0",[27,28,29,30,31,32,33,34,35,36],"agent","agent-skills","claude-code","cli","codex","devtools","evals","llm","opencode","testing","2026-08-18T22:49:47Z",false,"2026-08-11",[41],{"id":42,"name":43,"slug":44,"summary":45,"url":46,"kind":47,"platform":48,"author":49,"authorHandle":50,"publisher":51,"publishedAt":39,"about":52,"writtenBy":53},"entity_01kzsa7tgbextv2kt75xknj5mg","Over engineering a pipeline to email me about strangers’ skills for dynobox","over-engineering-a-pipeline-to-email-me-about-strangers-skills-for-dynobox","For the past few weeks I've been building Dynobox - a local test runner that records what an agent does inside a harness (Claude Code, Codex, OpenCode etc.) and lets you write deterministic assertions against its non-deterministic behavior: tool calls, shell commands, files created or changed and so on.","https:\u002F\u002Fx.com\u002Fbhkdotdev\u002Fstatus\u002F2087281766440546814","article","x","bhk","bhkdotdev","X",[],[],1786482227,{"id":56,"name":57,"slug":58,"description":59,"url":60,"githubUrl":60,"logoUrl":61,"openSource":17,"categories":62,"repo":64,"articles":75,"createdAt":76},"entity_01m0bp7xa7e4fbqpswv3j249vj","SkillForge","skillforge","A skill creator that proves its skills work. Evidence-driven skill creation for Claude Code and Codex: baseline-tested generation, per-skill regression evals, ecosystem doctor, cross-runtime compile, and an opt-in proactive advisor.","https:\u002F\u002Fgithub.com\u002Ftripleyak\u002FSkillForge","https:\u002F\u002Fr2.minima.ltd\u002Forg_01jakhtww5fk6b0jmxj0qfp9rf\u002Ftripleyak-8696ff1c.png",[63],{"name":5,"slug":6,"count":20},{"fullName":65,"url":60,"stars":66,"forks":67,"openIssues":20,"language":68,"license":69,"topics":70,"pushedAt":73,"archived":38,"licenseSpdx":69,"licenseOsi":17,"licenseVerifiedAt":74},"tripleyak\u002FSkillForge",856,88,"Python","MIT",[28,71,29,72,31,33],"claude-ai","claude-skills","2026-07-29T16:31:19Z","2026-08-19T00:20:26.791Z",[],1787098821,[78,82,86,90,94],{"name":79,"slug":80,"count":81},"Orchestration","orchestration",48,{"name":83,"slug":84,"count":85},"Harnesses","harnesses",38,{"name":87,"slug":88,"count":89},"Desktop Applications","desktop-applications",36,{"name":91,"slug":92,"count":93},"Skills & Prompts","skills-prompts",29,{"name":95,"slug":96,"count":97},"Frameworks & SDKs","frameworks-sdks",26]