Line of inquiry
Inquiring lines›How do system designs prevent auto…›What causes autonomous agents to e…›this line of inquiry
Which agent-created code should persist in infrastructure versus remain task-specific?
A broader line of inquiry — a family of 46 specific questions the research asks around this. Follow one into its inquiring-line page, or move sideways to a related line below.
Questions in this line of inquiry 46
Specific inquiring lines the field asks around this — ordered from the most general framing down to the most specific angle.
- Are durable shared code artifacts better than per-task harness patches?
- How do agents decide which created code deserves long-term persistence?
- Do agent improvements discovered on code tasks transfer to non-coding domains as well?
- How should agents decide which created code is worth persisting?
- How do agents decide which created code should persist versus disappear?
- Can skill libraries prevent redundant narrow artifacts from proliferating?
- How should harness infrastructure validate code that agents generate themselves?
- When should agent-created code be promoted into permanent harness infrastructure?
- How should AI skills be created and managed like software artifacts?
- How do agent-created code artifacts become part of harness infrastructure?
- What permission models govern code execution within agent skills?
- How should human oversight apply to persistent agent-authored code?
- Can skill validation through testing prevent unreliable programs from accumulating?
- Does codifying domain rules into agent scaffolding work at library scale?
- Can one-off agent code be safely promoted to durable infrastructure?
- What lifecycle management prevents in-loop skill creation from bloating an agent?
- What makes durable code artifacts more valuable than per-task harness patches?
- Can disposable agent-authored code be distinguished from reusable infrastructure?
- What design practices prevent tool manifests from varying across requests?
- What makes persistent, shared code artifacts from agents hard to manage at scale?
- When do agents benefit most from reusable workflow routines?
- How should skills be trusted and installed on sharing platforms?
- What makes a tool schema high-quality enough to prevent agent misuse?
- How do artifact families differ in matching verification scope to repair capability?
- Why treat tutorial videos as a separate supply line from agent trajectories?
- What metadata properties make code-derived skills auditable and comparable to their original source?
- What makes agent-initiated artifacts the underexplored frontier in harness engineering?
- How do skills authored in-loop validate faster than offline generated skills?
- Why do persistent, resynchronized artifacts compound harness capability gains?
- What makes skills suitable for retrieval and chaining in repositories?
- How do agents retrieve and compose skills from hierarchical multimodal wikis?
- What specific practices do developers use to prevent agent mistakes upfront?
- What distinguishes improvised spreadsheet workflows from institutionalized enterprise software?
- Can agents acquire new skills online when offline skill coverage runs out?
- How do agents discover and construct new APIs from existing applications?
- Can curator modules trained on one executor transfer to entirely different agent backbones?
- How does source-blind reconstruction verify that extracted skills are specific enough to be reusable?
- Did AIDE2's rewrites solve problems on a human checklist or search artifacts?
- How can we reorganize repositories to make behaviors easier to locate?
- Why does embedding research tools in coding assistants improve reliability?
- Does inspectable skill artifacts guarantee the behavior matches the person it claims to ground?
- How does Claude Code prevent git credentials from leaving the sandbox?
- How do skill libraries from human resources compare to hand-written skill libraries?
- How do diagnose-and-reshape loops compare to building new environments from scratch?
- What makes a distilled skill verifiable and ready for agent execution?
- How does the scoped credential inside Claude Code's sandbox differ from a full GitHub token?