pith. sign in
Pith Number

pith:RSLHGCOW

pith:2026:RSLHGCOWXB22X3YAG6SMUE27HA
not attested not anchored not stored refs resolved

EvoSkill: Automated Skill Discovery for Multi-Agent Systems

Jaydon Bingham, Noah Provenzano, Salaheddin Alzubi, Tu Vu, Weiyuan Chen

EvoSkill automatically discovers reusable agent skills through failure analysis to improve performance without changing the model.

arxiv:2603.02766 v1 · 2026-03-03 · cs.AI · cs.MA

Add to your LaTeX paper
\usepackage{pith}
\pithnumber{RSLHGCOWXB22X3YAG6SMUE27HA}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp
2 Internet Archive
3 Author claim open · sign in to claim
4 Citations open
5 Replications open
Portable graph bundle live · download bundle · merged state
The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

EvoSkill analyzes execution failures, proposes new skills or edits to existing ones, and materializes them into structured, reusable skill folders. A Pareto frontier of agent programs governs selection, retaining only skills that improve held-out validation performance while the underlying model remains frozen. We evaluate EvoSkill on two benchmarks: OfficeQA... improves exact-match accuracy by 7.3% (60.6% → 67.9%); and SealQA... yields a 12.1% gain (26.6% → 38.7%). ... skills evolved from SealQA transfers zero-shot to BrowseComp, improving accuracy by 5.3%.

C2weakest assumption

That iterative failure analysis can reliably generate skills whose benefits on held-out validation reflect genuine generalization rather than benchmark-specific artifacts or unstated implementation choices in the skill proposal step.

C3one line summary

EvoSkill evolves agent skills via failure analysis and Pareto frontier selection, raising exact-match accuracy 7.3% on OfficeQA and 12.1% on SealQA with 5.3% zero-shot transfer to BrowseComp.

References

55 extracted · 55 resolved · 6 Pith anchors

[1] Agent skills specification, 2025 2025
[2] GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning 2026 · arXiv:2507.19457
[3] Roma: Recursive open meta-agent framework for long-horizon multi-agent systems, 2026 2026
[4] Anthropic skills documentation, 2025 2025
[5] Claude code overview, 2026 2026

Formal links

2 machine-checked theorem links

Cited by

25 papers in Pith

Receipt and verification
First computed 2026-05-17T23:38:15.400988Z
Builder pith-number-builder-2026-05-17-v1
Signature Pith Ed25519 (pith-v1-2026-05) · public key
Schema pith-number/v1.0

Canonical hash

8c967309d6b875abef0037a4ca135f3812be2599ee232b74f19d79d336018128

Aliases

arxiv: 2603.02766 · arxiv_version: 2603.02766v1 · doi: 10.48550/arxiv.2603.02766 · pith_short_12: RSLHGCOWXB22 · pith_short_16: RSLHGCOWXB22X3YA · pith_short_8: RSLHGCOW
Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/RSLHGCOWXB22X3YAG6SMUE27HA \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 8c967309d6b875abef0037a4ca135f3812be2599ee232b74f19d79d336018128
Canonical record JSON
{
  "metadata": {
    "abstract_canon_sha256": "0a3216ff28967491494c1c732d2e6dfe23e008c046bcc7f01c5b06866da9d63e",
    "cross_cats_sorted": [
      "cs.MA"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.AI",
    "submitted_at": "2026-03-03T09:07:22Z",
    "title_canon_sha256": "d1ff511b1c5f7b42bcb3120baa9ae8d47138de7e880d195e6f7eb9e02db0d795"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2603.02766",
    "kind": "arxiv",
    "version": 1
  }
}