Pith Number

pith:EJC5MCXK

pith:2026:EJC5MCXKRJ522ASFLN62PFZFZN

not attested not anchored not stored refs resolved

Runtime-Structured Task Decomposition for Agentic Coding Systems

Bing Zhang, Chad DeLuca, Hima Patel, Ruchi Mahindru, Shubhi Asthana

Runtime-structured task decomposition reduces retry costs in agentic coding systems by rerunning only failed subtasks.

arxiv:2605.15425 v1 · 2026-05-14 · cs.SE · cs.AI

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{EJC5MCXKRJ522ASFLN62PFZFZN}

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

The runtime-structured approach reran only failed subtasks, reducing retry costs to 436 +/- 132 tokens for root cause analysis and 460 tokens for debugging, achieving up to 51.7% lower retry cost than monolithic systems and 73.2% lower retry cost than static decomposition baselines.

C2weakest assumption

That output validation against predefined schemas reliably catches errors and that subtask failures remain sufficiently localized to enable partial reruns without cascading effects or requiring full workflow restarts, as implied by the description of the runtime-structured configuration.

C3one line summary

Runtime-structured task decomposition reduces retry costs in agentic coding systems by up to 51.7% versus monolithic prompts by rerunning only failed subtasks on two software engineering workloads.

References

64 extracted · 64 resolved · 12 Pith anchors

[1] 30th USENIX Security Symposium (USENIX Security 21) , pages=

[2] Quantifying Memorization Across Neural Language Models · arXiv:2202.07646

[3] Differentiation

[4] AIDev: A Large-Scale Dataset of Real-World

[5] Panel: Privacy Challenges and Opportunities in \ LLM-Based \ Chatbot Applications , author=

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-20T00:00:57.958406Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

2245d60aea8a7bad02455b7da79725cb47ea36a3f005a69e07a5cbcc5e3bb03d

Aliases

arxiv: 2605.15425 · arxiv_version: 2605.15425v1 · doi: 10.48550/arxiv.2605.15425 · pith_short_12: EJC5MCXKRJ52 · pith_short_16: EJC5MCXKRJ522ASF · pith_short_8: EJC5MCXK

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/EJC5MCXKRJ522ASFLN62PFZFZN \
  | 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: 2245d60aea8a7bad02455b7da79725cb47ea36a3f005a69e07a5cbcc5e3bb03d

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "34904bc7e30d61671e47cbc0df63adbfb3a78dde9dd752611b83cc29453465c9",
    "cross_cats_sorted": [
      "cs.AI"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.SE",
    "submitted_at": "2026-05-14T21:16:23Z",
    "title_canon_sha256": "dd91e26b7ee12c4ab45d75686b4cebd75413159fac5765fb8d9aec86287218dc"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.15425",
    "kind": "arxiv",
    "version": 1
  }
}