# SkillProx: Self-Evolving Agent Skills via Proximal Textual Gradient Descent

Source: [arXiv](https://arxiv.org/abs/2608.07449v1)  
Feed7 permalink: https://feed7.dev/p/2608-07449v1-1co8km3  
Published: 2026-08-07T17:40:33.000Z  
Trust: Needs Review (needs_review)

## Why Included

SkillProx evolves reusable agent instructions through measured retries, regression rollback, and utility-based pruning, reporting a 3-point accuracy gain over its strongest baseline.

## Source Summary

SkillProx treats agent skills as editable text rather than model weights. Its **forward stage** reruns diagnosis-driven edits on the same task batch, rolls back regressions, and returns measured outcomes; its **backward stage** audits individual knowledge units for consolidation, demotion, or removal.

## Practical Implication

For recurring coding-agent work, the useful pattern is closed-loop skill maintenance: test each instruction change, preserve the prior version when results worsen, and periodically remove guidance that contributes no measured utility. This makes accumulated procedures more auditable than an ever-growing prompt file.

## Agent-Ready Context

SkillProx treats agent skills as editable text rather than model weights. Its **forward stage** reruns diagnosis-driven edits on the same task batch, rolls back regressions, and returns measured outcomes; its **backward stage** audits individual knowledge units for consolidation, demotion, or removal.

For recurring coding-agent work, the useful pattern is closed-loop skill maintenance: test each instruction change, preserve the prior version when results worsen, and periodically remove guidance that contributes no measured utility. This makes accumulated procedures more auditable than an ever-growing prompt file.

Across multiple backbones and in- and out-of-distribution benchmarks, SkillProx reports an average **3.0 percentage-point** accuracy gain over the strongest gradient-based baseline. The abstract does not identify the tasks, absolute scores, audit cost, or whether the approach transfers to live software repositories.

## Connected Context

Feed7 judgment across 409 accumulated Signals:

This operationalizes skill governance as reversible, measured maintenance: edit instructions, rerun the same tasks, roll back regressions, and periodically prune low-utility knowledge. It directly addresses the danger of endlessly accumulating procedures, but does not establish repository-scale practicality because task details, absolute results, audit cost, and live coding transfer are absent.

- [The Regression Tax: Decomposing Why Skills Help and Hurt LLM Agents](https://feed7.dev/p/2607-22520v1-0mz9wnf) — SkillProx turns the Regression Tax finding into a maintenance mechanism by explicitly detecting regressions, preserving prior versions, and auditing individual knowledge units instead of judging only aggregate gains.
- [Don't Ship Skills Without Evals — Philipp Schmid, Google DeepMind](https://feed7.dev/p/don-t-ship-skills-without-evals-philipp-schmid-google-deepmind-0fuh3ko) — The repeated with-and-without regression testing advocated here is a prerequisite for SkillProx’s rollback and pruning decisions; without harness-specific trials, its textual edits cannot be assigned measured utility reliably.
- [EveryInc/compound-engineering-plugin](https://feed7.dev/p/compound-engineering-plugin-06exf60) — Compound Engineering supplies the kind of persisted workflows and solved-problem notes SkillProx could maintain, while SkillProx adds a missing discipline for testing, reverting, consolidating, or removing that accumulated guidance.
- [WTF Is the Context Layer? The Missing Infrastructure for Production Agents — Prukalpa Sankar](https://feed7.dev/p/wtf-is-the-context-layer-the-missing-infrastructure-for-production-agent-0t47xqf) — Atlan’s shared, versioned context broadens the storage and portability layer for skills, while SkillProx contributes a candidate method for deciding which versioned instructions should be promoted, demoted, or removed.

## Context Map

- Layer: agent
- Domains: coding
- Topics: skills, agent-memory, agent-evals

## Uncertainty

- Across multiple backbones and in- and out-of-distribution benchmarks, SkillProx reports an average **3.0 percentage-point** accuracy gain over the strongest gradient-based baseline. The abstract does not identify the tasks, absolute scores, audit cost, or whether the approach transfers to live software repositories.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.