Results for “meta-harness”

5 skills

self-improvement-loops

Designs and governs recursive self-improvement loops where an agent mines its own failures and proposes edits to its own harness, prompts, or workflow, covering acceptance gates, diversity preservation, and the optimization ladder.

16.9k · bundle

More results

harness-engineering

Prevent repeated AI coding-agent mistakes by turning failures into durable instructions, drift checks, regression tests, failure memory, and adoption reports tailored to the target repository.

harness-engineering

Designs autonomous agent harnesses with locked evaluators, editable surfaces, durable logging, novelty gates, pruning, rollback, and human approval boundaries.

gan-style-harness

Uses a multi-agent generator-evaluator feedback loop to build high-quality applications from a single prompt, inspired by GANs and Anthropic's harness design.

orchestra-research

transformer-lens-interpretability

Inspect and manipulate transformer internals via HookPoints and activation caching for mechanistic interpretability research.

10.4k · bundle