{"paper":{"title":"Before the First Edit? Measuring When Coding-Agent Guidance Actually Arrives","arxiv_id":"2610.35992","primary_cat":"cs.SE","submitted_at":"2026-10-26T12:22:34+00:00","version_compiled":"v1"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v1","sentences":[{"i":0,"p":0.55,"id":"s0","own":false,"role":"gap","text":"When a coding-agent harness promises guidance before the first source edit, what event does it actually intercept?","hedge":{"p":0.98,"label":"unclear"},"own_p":0.06,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.55,"method":0.0,"result":0.0,"background":0.45,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.81,"id":"s1","own":true,"role":"method","text":"We audit five separate events: policy assignment, request eligibility, trigger, plan delivery, and observed source-content change.","hedge":{"p":0.84,"label":"asserted"},"own_p":0.91,"main_claim_p":0.03,"quantitative":true,"quantitative_p":0.81,"role_probabilities":{"gap":0.0,"method":0.81,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.19,"interpretation":0.0}},{"i":2,"p":0.55,"id":"s2","own":true,"role":"gap","text":"A prescribed-action comparison exposes the coverage of three gate rules but cannot tell us how an agent would continue after receiving a plan.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.7,"main_claim_p":0.06,"quantitative":true,"quantitative_p":0.77,"role_probabilities":{"gap":0.55,"method":0.02,"result":0.01,"background":0.0,"limitation":0.41,"future_work":0.0,"contribution":0.0,"interpretation":0.01}},{"i":3,"p":0.99,"id":"s3","own":true,"role":"method","text":"We therefore run local agents on public software issues and hash Python files before and after each executed action.","hedge":{"p":0.91,"label":"asserted"},"own_p":0.96,"main_claim_p":0.03,"quantitative":false,"quantitative_p":0.06,"role_probabilities":{"gap":0.0,"method":0.99,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.01,"interpretation":0.0}},{"i":4,"p":0.9,"id":"s4","own":true,"role":"method","text":"In a pilot-guided, 76-arm study using Qwen2.5-Coder 14B, 15 arms change Python source across seven tasks and three repositories.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.95,"main_claim_p":0.05,"quantitative":true,"quantitative_p":0.99,"role_probabilities":{"gap":0.0,"method":0.9,"result":0.01,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.09,"interpretation":0.0}},{"i":5,"p":0.93,"id":"s5","own":true,"role":"result","text":"Guidance precedes every observed first change, but every change uses a structured editor and the study misses its prespecified 20-arm coverage target.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.95,"main_claim_p":0.29,"quantitative":true,"quantitative_p":0.96,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.93,"background":0.0,"limitation":0.07,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":6,"p":1.0,"id":"s6","own":true,"role":"result","text":"Post hoc local tests find seven changed arms that pass the tested target and regression selectors; four of those seven use incomplete regression lists.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.95,"main_claim_p":0.09,"quantitative":true,"quantitative_p":0.98,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":7,"p":0.47,"id":"s7","own":false,"role":"limitation","text":"Public PAUSE and SI2CA trajectories add source-text changes bracketed by adjacent reads, but lack synchronized filesystem states and cannot identify exact first mutations.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.14,"main_claim_p":0.05,"quantitative":false,"quantitative_p":0.35,"role_probabilities":{"gap":0.15,"method":0.03,"result":0.06,"background":0.28,"limitation":0.47,"future_work":0.0,"contribution":0.0,"interpretation":0.01}},{"i":8,"p":0.52,"id":"s8","own":true,"role":"interpretation","text":"The audit makes the timing claim testable within a stated source scope.","hedge":{"p":0.88,"label":"asserted"},"own_p":0.88,"main_claim_p":0.14,"quantitative":false,"quantitative_p":0.05,"role_probabilities":{"gap":0.0,"method":0.01,"result":0.02,"background":0.0,"limitation":0.2,"future_work":0.0,"contribution":0.25,"interpretation":0.52}},{"i":9,"p":0.73,"id":"s9","own":true,"role":"limitation","text":"These data do not show that planning improves repair correctness or covers shell-first writes in live agent runs.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.81,"main_claim_p":0.6,"quantitative":false,"quantitative_p":0.07,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.07,"background":0.0,"limitation":0.73,"future_work":0.0,"contribution":0.0,"interpretation":0.2}}]},"compiled":{"main_claim":{"p":0.81,"text":"These data do not show that planning improves repair correctness or covers shell-first writes in live agent runs.","location":"abstract s9","hedge_strength":{"p":0.98,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":7416,"cost_usd_reported":0.0},"compiled_at":"2026-10-28","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.06,"states_limitation":0.98,"human_participants":0.05,"controlled_comparison":0.17,"randomized_assignment":0.07,"states_quantitative_result":0.97,"compares_with_prior_results":0.63,"mentions_code_or_data_availability":0.06},"evidential_basis":{"p":1.0,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.0,"analysis_of_existing_data":0.0,"observation_or_experiment":1.0,"simulation_or_computation":0.0}},"contribution_type":{"p":1.0,"value":"empirical study","probabilities":{"review":0.0,"method_or_tool":0.0,"empirical_study":1.0,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":0.86,"text":"We therefore run local agents on public software issues and hash Python files before and after each executed action.","location":"abstract s3"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v1"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":10,"sentences_in_abstract":10},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}