Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note2 min read

Proof-or-Stop: Don't Trust the Agent, Trust the Evidence -- Loop Engineering for Verifiable Evidence-Gated Lifecycle Control

Source

arXiv

arxiv.org/abs/2607.14890v1

Read the source

Source

Plain-Language Brief

The paper presents Proof-or-Stop Lifecycle Control, a method that ensures software work transitions only when supported by verifiable evidence. The method was tested using an open-source implementation and showed promising results in reducing false-DONE rates and amplification of issues. However, the evaluation is limited to one model family and a self-hosted corpus.

What the Document Says

  • The paper introduces Proof-or-Stop Lifecycle Control for autonomous coding agents.
  • It allows lifecycle transitions only when fresh, tracked-source-state-bound, mechanically verifiable evidence satisfies the relevant gate.
  • Proof is treated as gate-admissible evidence under a stated trust model, not semantic program correctness.
  • The method was tested using an open-source implementation.
  • The evaluation includes mechanism tests, powered control-policy ablation, and operated self-application evidence.
  • Results show the unattended-loop engine passed 10 of 10 scenarios with zero false-DONE and rejected 18 tamper classes with zero false accepts.
  • In a 9,240-cell ablation, the pre-registered A4 versus A2-prime comparison reduced visible-pass/hidden-fail amplification from 31 of 1,800 injected cells to 2 of 1,800 under the gated loop.
  • The self-application corpus contains 565 stories and 1,007 review findings, with 94.8% resolved.
  • The evaluation is limited to one model family, 24 ablation tasks, and a self-hosted corpus.

Key Concepts Mentioned

  • autonomous coding agents
  • lifecycle control
  • proof-or-stop
  • verifiable evidence
  • evaluation

Claims Made by the Document

  • Proof-or-Stop Lifecycle Control reduces false-DONE rates and amplification of issues.

Evidence Used

  • Results from mechanism tests, powered control-policy ablation, and operated self-application evidence.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Autonomous coding agents increasingly execute multi-step software work, but lifecycle states such as reviewed, tested, DONE, and ready-to-merge remain claims unless supported by current evidence. We present Proof-or-Stop Lifecycle Control, a method that permits lifecycle transitions only when fresh, tracked-source-state-bound, mechanically verifiable evidence satisfies the relevant gate. The method treats agent outputs as claims rather than lifecycle state, and uses proof operationally to mean gate-admissible evidence under a stated trust model, not semantic program correctness. We evaluate an open-source implementation through mechanism tests, a powered control-policy ablation, and operated self-application evidence. The unattended-loop engine passed 10 of 10 scenarios with zero false-DONE, and local-key receipt bundles rejected 18 tamper classes with zero false accepts. In a 9,240-cell ablation, the pre-registered A4 versus A2-prime comparison reduced visible-pass/hidden-fail amplification from 31 of 1,800 injected cells under a compute-budgeted naive loop to 2 of 1,800 under the gated loop, a 1.6 percentage-point improvement in not-amplified rate with a 95 percent confidence interval of [0.8, 2.5]. A near-compute A3 versus A4 comparison, 14 of 1,800 versus 2 of 1,800, indicates that the gain is associated with enforcing review as a lifecycle gate rather than merely adding a reviewer. The self-application corpus contains 565 stories and 1,007 review findings, with 94.8 percent resolved, plus a 68-row high/critical cross-vendor exhibit. These results support Proof-or-Stop as a model-agnostic, host-neutral control layer for deciding which autonomous-agent claims a lifecycle may act on. The evaluation is limited to one model family, 24 ablation tasks, and a self-hosted corpus.