{
  "schema_version": "1.1",
  "id": "archive:https://arxiv.org/abs/2608.07449v1",
  "slug": "2608-07449v1-1co8km3",
  "url": "https://feed7.dev/p/2608-07449v1-1co8km3",
  "title": "SkillProx: Self-Evolving Agent Skills via Proximal Textual Gradient Descent",
  "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.",
  "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_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.\n\nFor 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.\n\nAcross 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.",
  "source": {
    "name": "arXiv",
    "url": "https://arxiv.org/abs/2608.07449v1",
    "published_at": "2026-08-07T17:40:33.000Z"
  },
  "source_class": "blog_post",
  "content_type": "Paper",
  "layer": "agent",
  "domains": [
    "coding"
  ],
  "topics": [
    "skills",
    "agent-memory",
    "agent-evals"
  ],
  "verification": {
    "status": "needs_review",
    "label": "Needs Review",
    "method": "unverified",
    "verified_at": null
  },
  "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."
  ],
  "connected_context": {
    "meaning": "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.",
    "corpus_size": 409,
    "generated_at": "2026-08-10T10:05:29.932Z",
    "connections": [
      {
        "title": "The Regression Tax: Decomposing Why Skills Help and Hurt LLM Agents",
        "source_name": "arXiv",
        "source_url": "https://arxiv.org/abs/2607.22520v1",
        "feed7_url": "https://feed7.dev/p/2607-22520v1-0mz9wnf",
        "reason": "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."
      },
      {
        "title": "Don't Ship Skills Without Evals — Philipp Schmid, Google DeepMind",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=0vphxNt4wyk",
        "feed7_url": "https://feed7.dev/p/don-t-ship-skills-without-evals-philipp-schmid-google-deepmind-0fuh3ko",
        "reason": "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."
      },
      {
        "title": "EveryInc/compound-engineering-plugin",
        "source_name": "GitHub",
        "source_url": "https://github.com/EveryInc/compound-engineering-plugin",
        "feed7_url": "https://feed7.dev/p/compound-engineering-plugin-06exf60",
        "reason": "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."
      },
      {
        "title": "WTF Is the Context Layer? The Missing Infrastructure for Production Agents — Prukalpa Sankar",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=8G_1-3IO4ZQ",
        "feed7_url": "https://feed7.dev/p/wtf-is-the-context-layer-the-missing-infrastructure-for-production-agent-0t47xqf",
        "reason": "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."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-08-07T17:40:33.000Z",
  "modified_at": "2026-08-07T17:40:33.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/2608-07449v1-1co8km3",
    "json": "https://feed7.dev/p/2608-07449v1-1co8km3.json",
    "markdown": "https://feed7.dev/p/2608-07449v1-1co8km3.md"
  }
}