loopmath 0.3.3__tar.gz → 0.3.5__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (444) hide show
  1. {loopmath-0.3.3 → loopmath-0.3.5}/PKG-INFO +19 -16
  2. {loopmath-0.3.3 → loopmath-0.3.5}/README.md +18 -15
  3. {loopmath-0.3.3 → loopmath-0.3.5}/pyproject.toml +2 -2
  4. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/__init__.py +1 -1
  5. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/cli.py +19 -3
  6. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/cli_registry.py +20 -8
  7. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/onboard/commands.py +1 -1
  8. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/prices.toml +29 -3
  9. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/skill/commands.py +17 -19
  10. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/skill/doctor.py +35 -8
  11. loopmath-0.3.5/src/loopmath/skill/guide.py +52 -0
  12. loopmath-0.3.3/src/loopmath/skill/skills/loopmath-update-fit/SKILL.md → loopmath-0.3.5/src/loopmath/skill/guides/fit.md +2 -8
  13. loopmath-0.3.3/src/loopmath/skill/skills/loopmath-import-runs/SKILL.md → loopmath-0.3.5/src/loopmath/skill/guides/import.md +1 -7
  14. loopmath-0.3.3/src/loopmath/skill/skills/loopmath-onboard/SKILL.md → loopmath-0.3.5/src/loopmath/skill/guides/onboard.md +2 -8
  15. loopmath-0.3.3/src/loopmath/skill/skills/loopmath-plan-task/SKILL.md → loopmath-0.3.5/src/loopmath/skill/guides/plan.md +3 -9
  16. loopmath-0.3.3/src/loopmath/skill/skills/loopmath-record-run/SKILL.md → loopmath-0.3.5/src/loopmath/skill/guides/record.md +2 -8
  17. {loopmath-0.3.3/src/loopmath/skill/skills → loopmath-0.3.5/src/loopmath/skill/guides}/reference.md +3 -3
  18. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/skill/install.py +144 -71
  19. loopmath-0.3.5/src/loopmath/skill/skills/loopmath/SKILL.md +57 -0
  20. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/commands.py +2 -0
  21. loopmath-0.3.5/src/loopmath/update.py +299 -0
  22. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath.egg-info/PKG-INFO +19 -16
  23. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath.egg-info/SOURCES.txt +9 -6
  24. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_price.py +5 -4
  25. loopmath-0.3.5/tests/test_update_check.py +428 -0
  26. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_v027_small_fixes.py +1 -1
  27. loopmath-0.3.3/src/loopmath/skill/skills/loopmath/SKILL.md +0 -35
  28. {loopmath-0.3.3 → loopmath-0.3.5}/LICENSE +0 -0
  29. {loopmath-0.3.3 → loopmath-0.3.5}/setup.cfg +0 -0
  30. {loopmath-0.3.3 → loopmath-0.3.5}/spec/ocp-v0.2.schema.json +0 -0
  31. {loopmath-0.3.3 → loopmath-0.3.5}/spec/ocp-v0.schema.json +0 -0
  32. {loopmath-0.3.3 → loopmath-0.3.5}/spec/ocp_conformance.py +0 -0
  33. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/__main__.py +0 -0
  34. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapter_host.py +0 -0
  35. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/__init__.py +0 -0
  36. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/atrium.py +0 -0
  37. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/base.py +0 -0
  38. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/bb.py +0 -0
  39. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/bb_support.py +0 -0
  40. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/evolve.py +0 -0
  41. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/harness2.py +0 -0
  42. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/omp.py +0 -0
  43. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/omp_facts.py +0 -0
  44. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/omp_links.py +0 -0
  45. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/omp_parse.py +0 -0
  46. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/omp_types.py +0 -0
  47. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/opencode.py +0 -0
  48. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/opencode_store.py +0 -0
  49. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/openevolve.py +0 -0
  50. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/orca.py +0 -0
  51. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/otel_genai.py +0 -0
  52. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/otel_genai_graph.py +0 -0
  53. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/otel_genai_parse.py +0 -0
  54. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/otel_genai_types.py +0 -0
  55. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/otel_genai_usage.py +0 -0
  56. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/paseo.py +0 -0
  57. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/pi.py +0 -0
  58. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/pi_support.py +0 -0
  59. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/plugin_eval.py +0 -0
  60. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/registry.py +0 -0
  61. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/adapters/shinkaevolve.py +0 -0
  62. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/__init__.py +0 -0
  63. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/block.py +0 -0
  64. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/check_pymc.py +0 -0
  65. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/commands.py +0 -0
  66. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/compose.py +0 -0
  67. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/design.py +0 -0
  68. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/features.py +0 -0
  69. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/fit.py +0 -0
  70. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/forest.py +0 -0
  71. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/gaussian.py +0 -0
  72. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/integrated.py +0 -0
  73. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/lookahead.py +0 -0
  74. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/outcome.py +0 -0
  75. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/pricing.py +0 -0
  76. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/priors.py +0 -0
  77. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/rules.py +0 -0
  78. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/sampled.py +0 -0
  79. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/belief/state.py +0 -0
  80. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/builder/__init__.py +0 -0
  81. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/builder/cli.py +0 -0
  82. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/builder/context.py +0 -0
  83. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/builder/predict.py +0 -0
  84. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/builder/server.py +0 -0
  85. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/cli_errors.py +0 -0
  86. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/cli_graph.py +0 -0
  87. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/cli_support.py +0 -0
  88. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/__init__.py +0 -0
  89. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/_thresholds.py +0 -0
  90. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/_thresholds_sw.py +0 -0
  91. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/calibrate.py +0 -0
  92. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/calibration.py +0 -0
  93. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/checks.py +0 -0
  94. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/commands.py +0 -0
  95. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/model.py +0 -0
  96. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/units.py +0 -0
  97. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/compare/value.py +0 -0
  98. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/__init__.py +0 -0
  99. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/arms.py +0 -0
  100. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/cli_e0.py +0 -0
  101. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/estimate.py +0 -0
  102. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/estimate_rework.py +0 -0
  103. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/figures.py +0 -0
  104. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/io.py +0 -0
  105. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/e0/verb.py +0 -0
  106. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/explain.py +0 -0
  107. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/fit.py +0 -0
  108. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/fit_assembly.py +0 -0
  109. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/fit_bayes.py +0 -0
  110. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/fit_masks.py +0 -0
  111. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/fit_predict.py +0 -0
  112. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/fit_pricing.py +0 -0
  113. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/fit_transfer.py +0 -0
  114. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gaps.py +0 -0
  115. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gate/__init__.py +0 -0
  116. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gate/commands.py +0 -0
  117. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gate/core.py +0 -0
  118. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gate/inputs.py +0 -0
  119. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gate/winner.py +0 -0
  120. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gate/winner_cli.py +0 -0
  121. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gate/winner_model.py +0 -0
  122. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/gitwalk.py +0 -0
  123. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/grade.py +0 -0
  124. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/grade_support.py +0 -0
  125. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/__init__.py +0 -0
  126. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/artifact_git.py +0 -0
  127. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/artifact_join.py +0 -0
  128. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/artifact_kinds.py +0 -0
  129. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/artifacts.py +0 -0
  130. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/bashparse.py +0 -0
  131. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/bashread_ops.py +0 -0
  132. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/bashwrite_ops.py +0 -0
  133. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/bashwrites.py +0 -0
  134. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/cache.py +0 -0
  135. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/codexio.py +0 -0
  136. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/codexio_parse.py +0 -0
  137. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/codexio_scan.py +0 -0
  138. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/dataset.py +0 -0
  139. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/dataset_build.py +0 -0
  140. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/dataset_core.py +0 -0
  141. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/dataset_edges.py +0 -0
  142. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/dataset_nodes.py +0 -0
  143. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/extract.py +0 -0
  144. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/html_common.py +0 -0
  145. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/html_costcurve.py +0 -0
  146. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/html_css.py +0 -0
  147. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/html_data.py +0 -0
  148. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/html_force.py +0 -0
  149. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/html_render.py +0 -0
  150. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/html_swimlanes.py +0 -0
  151. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler.py +0 -0
  152. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_cli.py +0 -0
  153. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_common.py +0 -0
  154. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_parse.py +0 -0
  155. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_pricing.py +0 -0
  156. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt.py +0 -0
  157. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_build.py +0 -0
  158. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_dispatch.py +0 -0
  159. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_evidence.py +0 -0
  160. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_node.py +0 -0
  161. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_records.py +0 -0
  162. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_result.py +0 -0
  163. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_session.py +0 -0
  164. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_tools.py +0 -0
  165. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_validation.py +0 -0
  166. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_prompt_workflow.py +0 -0
  167. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_report.py +0 -0
  168. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labeler_score.py +0 -0
  169. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/labels.py +0 -0
  170. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/launch.py +0 -0
  171. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/launch_origin.py +0 -0
  172. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/ocp.py +0 -0
  173. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/ocp_emit.py +0 -0
  174. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/ocp_support.py +0 -0
  175. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/render.py +0 -0
  176. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/runfile.py +0 -0
  177. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/runfile_export.py +0 -0
  178. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/runfile_finish.py +0 -0
  179. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/scan.py +0 -0
  180. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/schema.py +0 -0
  181. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/graph/token_completeness.py +0 -0
  182. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness/__init__.py +0 -0
  183. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness/layouts.py +0 -0
  184. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness_runs/__init__.py +0 -0
  185. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness_runs/commands.py +0 -0
  186. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness_runs/costs.py +0 -0
  187. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness_runs/ocp.py +0 -0
  188. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness_runs/rows.py +0 -0
  189. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/harness_runs/streams.py +0 -0
  190. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/__init__.py +0 -0
  191. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/base.py +0 -0
  192. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/claude_code.py +0 -0
  193. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/claude_code_support.py +0 -0
  194. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/codex.py +0 -0
  195. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/codex_support.py +0 -0
  196. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/ocp.py +0 -0
  197. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/ocp_analysis.py +0 -0
  198. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/ocp_artifact.py +0 -0
  199. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/ocp_common.py +0 -0
  200. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/ocp_convert.py +0 -0
  201. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ingest/ocp_projection.py +0 -0
  202. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/integrations/__init__.py +0 -0
  203. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/integrations/dspy.py +0 -0
  204. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/__init__.py +0 -0
  205. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/artifacts.py +0 -0
  206. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/clip.py +0 -0
  207. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/codex_window.py +0 -0
  208. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/costs.py +0 -0
  209. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/match.py +0 -0
  210. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/settle.py +0 -0
  211. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/logmatch/tariff.py +0 -0
  212. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/__init__.py +0 -0
  213. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/canonical.py +0 -0
  214. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/commands.py +0 -0
  215. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/conformance.py +0 -0
  216. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/contractv3.py +0 -0
  217. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/emit.py +0 -0
  218. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/graph_contract.py +0 -0
  219. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/membership.py +0 -0
  220. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/migrate.py +0 -0
  221. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/schema/ocp-v0.2.schema.json +0 -0
  222. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/schema/ocp-v0.3.schema.json +0 -0
  223. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/schema/ocp-v0.schema.json +0 -0
  224. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/ocp/version.py +0 -0
  225. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/onboard/__init__.py +0 -0
  226. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/onboard/history.py +0 -0
  227. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/onboard/label.py +0 -0
  228. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/onboard/usual.py +0 -0
  229. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/output.py +0 -0
  230. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/pool.py +0 -0
  231. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/price.py +0 -0
  232. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/price_validation.py +0 -0
  233. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/__init__.py +0 -0
  234. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/benchmarks.py +0 -0
  235. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/benchmarks.toml +0 -0
  236. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/build.py +0 -0
  237. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/bundle/README.md +0 -0
  238. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/bundle/e0.jsonl.gz +0 -0
  239. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/bundle/lanes.jsonl.gz +0 -0
  240. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/bundle/manifest.json +0 -0
  241. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/bundle/rq1.jsonl.gz +0 -0
  242. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/bundle/sweep.jsonl.gz +0 -0
  243. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/commands.py +0 -0
  244. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/e0.py +0 -0
  245. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/history.py +0 -0
  246. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/lanes.py +0 -0
  247. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/ocpdoc.py +0 -0
  248. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/reduce.py +0 -0
  249. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/registry.py +0 -0
  250. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/rq1.py +0 -0
  251. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/show.py +0 -0
  252. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/sweep.py +0 -0
  253. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/priors/validate.py +0 -0
  254. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/receipts.py +0 -0
  255. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/__init__.py +0 -0
  256. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/backlog.py +0 -0
  257. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/commands.py +0 -0
  258. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/curve.py +0 -0
  259. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/engine.py +0 -0
  260. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/gain.py +0 -0
  261. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/message.py +0 -0
  262. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/receipts.py +0 -0
  263. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/search.py +0 -0
  264. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/stats.py +0 -0
  265. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/storeread.py +0 -0
  266. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend/variants.py +0 -0
  267. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/recommend_role.py +0 -0
  268. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/report/__init__.py +0 -0
  269. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/report/eligibility.py +0 -0
  270. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/report/html.py +0 -0
  271. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/report/terminal.py +0 -0
  272. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/report/terminal_exclusions.py +0 -0
  273. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/report/terminal_format.py +0 -0
  274. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/report/terminal_read.py +0 -0
  275. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/research_defaults.py +0 -0
  276. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/research_paths.py +0 -0
  277. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/scoring.py +0 -0
  278. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/scoring_comparisons.py +0 -0
  279. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/scoring_predictive.py +0 -0
  280. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/screen/__init__.py +0 -0
  281. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/screen/commands.py +0 -0
  282. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/screen/design.py +0 -0
  283. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/screen/edits.py +0 -0
  284. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/screen/model.py +0 -0
  285. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/screen/record.py +0 -0
  286. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/share/__init__.py +0 -0
  287. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/share/commands.py +0 -0
  288. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/share/export.py +0 -0
  289. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/share/import_.py +0 -0
  290. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/skill/__init__.py +0 -0
  291. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/__init__.py +0 -0
  292. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/budget.py +0 -0
  293. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/config.py +0 -0
  294. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/finish.py +0 -0
  295. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/fitjob.py +0 -0
  296. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/home.py +0 -0
  297. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/ids.py +0 -0
  298. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/lock.py +0 -0
  299. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/record.py +0 -0
  300. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/report.py +0 -0
  301. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/rule_change.py +0 -0
  302. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/run_names.py +0 -0
  303. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/runs.py +0 -0
  304. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/status.py +0 -0
  305. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/store/verify.py +0 -0
  306. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/surface.py +0 -0
  307. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/surface_bootstrap.py +0 -0
  308. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/surface_estimation.py +0 -0
  309. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/surface_frame.py +0 -0
  310. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/surface_presentation.py +0 -0
  311. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/surface_support.py +0 -0
  312. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/taskmodel.py +0 -0
  313. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/tasks/__init__.py +0 -0
  314. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/tasks/commands.py +0 -0
  315. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/tasks/data.py +0 -0
  316. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/tasks/fit.py +0 -0
  317. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/tasks/posterior.py +0 -0
  318. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/tasks/select.py +0 -0
  319. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/tasks/sim.py +0 -0
  320. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/types.py +0 -0
  321. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/__init__.py +0 -0
  322. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/README.md +0 -0
  323. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/base.css +0 -0
  324. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/builder.css +0 -0
  325. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/builder.html +0 -0
  326. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/builder.js +0 -0
  327. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/common.js +0 -0
  328. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/estimates.css +0 -0
  329. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/estimates.js +0 -0
  330. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/graph.css +0 -0
  331. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/graph.js +0 -0
  332. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/plans.css +0 -0
  333. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/plans.js +0 -0
  334. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/results.css +0 -0
  335. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/results.js +0 -0
  336. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/runs.js +0 -0
  337. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/viz.css +0 -0
  338. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/assets/viz.js +0 -0
  339. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/common.py +0 -0
  340. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/plans.py +0 -0
  341. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/posterior.py +0 -0
  342. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/views/runs.py +0 -0
  343. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/__init__.py +0 -0
  344. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/candidates.py +0 -0
  345. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/catalog/best_of_n.toml +0 -0
  346. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/catalog/implement_review.toml +0 -0
  347. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/catalog/plan_implement.toml +0 -0
  348. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/catalog/plan_implement_review.toml +0 -0
  349. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/catalog/solo.toml +0 -0
  350. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/catalog/swarm.toml +0 -0
  351. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/commands.py +0 -0
  352. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/diff.py +0 -0
  353. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/format.py +0 -0
  354. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/graphview.py +0 -0
  355. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/ids.py +0 -0
  356. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/infer.py +0 -0
  357. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/models.py +0 -0
  358. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/ocp.py +0 -0
  359. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath/workflows/shapes.py +0 -0
  360. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath.egg-info/dependency_links.txt +0 -0
  361. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath.egg-info/entry_points.txt +0 -0
  362. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath.egg-info/requires.txt +0 -0
  363. {loopmath-0.3.3 → loopmath-0.3.5}/src/loopmath.egg-info/top_level.txt +0 -0
  364. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_atrium.py +0 -0
  365. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_bb.py +0 -0
  366. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_harness2.py +0 -0
  367. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_omp.py +0 -0
  368. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_opencode.py +0 -0
  369. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_opencode_output_tokens.py +0 -0
  370. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_orca.py +0 -0
  371. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_otel_genai.py +0 -0
  372. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_paseo.py +0 -0
  373. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapter_services.py +0 -0
  374. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapters.py +0 -0
  375. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_adapters_pi.py +0 -0
  376. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_analyze_cli.py +0 -0
  377. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_arms.py +0 -0
  378. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_cli_graph.py +0 -0
  379. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_cli_help_lists.py +0 -0
  380. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_cli_labeler.py +0 -0
  381. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_cli_main_module.py +0 -0
  382. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_cli_plan_hidden.py +0 -0
  383. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_cli_prices.py +0 -0
  384. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_cli_startup.py +0 -0
  385. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_color_proof.py +0 -0
  386. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_duplicate_guards.py +0 -0
  387. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_estimate.py +0 -0
  388. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_explain.py +0 -0
  389. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_fit.py +0 -0
  390. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_gaps.py +0 -0
  391. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_gitwalk.py +0 -0
  392. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_grade.py +0 -0
  393. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_artifact_recording.py +0 -0
  394. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_artifacts.py +0 -0
  395. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_bashwrites.py +0 -0
  396. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_cache.py +0 -0
  397. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_codexio.py +0 -0
  398. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_dataset.py +0 -0
  399. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_extract.py +0 -0
  400. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_glue.py +0 -0
  401. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_html.py +0 -0
  402. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_labeler.py +0 -0
  403. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_labeler_evidence_matrix.py +0 -0
  404. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_launch.py +0 -0
  405. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_ocp.py +0 -0
  406. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_render.py +0 -0
  407. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_runfile.py +0 -0
  408. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_graph_schema.py +0 -0
  409. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_home_guard.py +0 -0
  410. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_html.py +0 -0
  411. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ingest.py +0 -0
  412. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ingest_claude_code.py +0 -0
  413. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ingest_codex.py +0 -0
  414. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ingest_ocp.py +0 -0
  415. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_io.py +0 -0
  416. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_not_run_summary.py +0 -0
  417. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ocp_conformance.py +0 -0
  418. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ocp_golden.py +0 -0
  419. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ocp_open_enums.py +0 -0
  420. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_ocp_versioning.py +0 -0
  421. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_output_html_path.py +0 -0
  422. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_packaging_metadata.py +0 -0
  423. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_packaging_smoke.py +0 -0
  424. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_pool.py +0 -0
  425. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_preview_script.py +0 -0
  426. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_receipts.py +0 -0
  427. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_release_paths.py +0 -0
  428. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_report_eligibility.py +0 -0
  429. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_research_cli.py +0 -0
  430. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_runfile_privacy_n1b.py +0 -0
  431. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_scoring.py +0 -0
  432. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_split_rebindings.py +0 -0
  433. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_starting_prior.py +0 -0
  434. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_surface.py +0 -0
  435. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_taskmodel_features.py +0 -0
  436. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_terminal.py +0 -0
  437. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_types_control.py +0 -0
  438. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_types_piece_prediction.py +0 -0
  439. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_v01_scaffold.py +0 -0
  440. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_v027_p3.py +0 -0
  441. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_v028_sol_6_1.py +0 -0
  442. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_v031_help_labels.py +0 -0
  443. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_v031_json_errors_and_costs.py +0 -0
  444. {loopmath-0.3.3 → loopmath-0.3.5}/tests/test_v031_new_findings.py +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: loopmath
3
- Version: 0.3.3
3
+ Version: 0.3.5
4
4
  Summary: Plan agent workflows, record how they went, and learn from your own Claude Code and Codex logs
5
5
  Author: loopmath
6
6
  License-Expression: MIT
@@ -42,7 +42,7 @@ loopmath helps an orchestrator agent choose how to run a coding task, and learns
42
42
 
43
43
  Before a task, it predicts how the user's usual agent workflow will do: the chance of an accepted result, and the cost in tokens and dollars. It shows the other workflows on a success-cost curve and offers two exploration picks. Each pick comes with its gain on every future similar run, its one-time price, and the number of runs it takes to pay back. After the task, the orchestrator records what ran and what came of it. loopmath fills in the token costs from the Claude Code and Codex logs, writes a receipt of predicted against actual, and refits.
44
44
 
45
- loopmath never starts or coordinates agents. Claude Code, Codex, herdr or your own tool does that; loopmath is a CLI they call. Everything runs locally. loopmath connects to no other machine: `loopmath builder` serves its page on 127.0.0.1 only, and no other command opens a network connection. The logs are opened read only, and loopmath writes only to its store (`~/.loopmath`, which holds its caches too), to the skill folders `skill install` names, and to paths you name.
45
+ loopmath never starts or coordinates agents. Claude Code, Codex, herdr or your own tool does that; loopmath is a CLI they call. Everything runs locally. loopmath connects to no other machine but pypi.org: `loopmath builder` serves its page on 127.0.0.1 only, and no other command opens a network connection, except the daily update check. Once a day loopmath asks pypi.org whether a newer loopmath exists; it sends only its version and uploads nothing. `LOOPMATH_NO_UPDATE_CHECK=1` turns the check off, and it is off when `CI` is set. The logs are opened read only, and loopmath writes only to its store (`~/.loopmath`, which holds its caches too), to the skill folders `skill install` names, and to paths you name.
46
46
 
47
47
  The package installs two commands: `loopmath` and its short alias `loop`. If another `loop` is already on your PATH, use `loopmath`.
48
48
 
@@ -64,6 +64,8 @@ python3 -m venv .venv
64
64
 
65
65
  Installing needs the network: pip, pipx and uv download the build backend (setuptools 77 or newer) and the dependencies (jsonschema, numpy, pandas, scipy), from a clone too, so `pip install --no-index -e .` fails offline.
66
66
 
67
+ To upgrade, `loopmath update` says whether a newer loopmath is out and prints the command for how you installed it (`uv tool upgrade loopmath`, `pipx upgrade loopmath`, or pip); run it, then `loopmath skill install`. loopmath never upgrades itself.
68
+
67
69
  ## First run
68
70
 
69
71
  ```sh
@@ -73,20 +75,21 @@ loopmath skill install # teach your coding agents how to use loopmat
73
75
 
74
76
  Then ask your agent to onboard, for example "onboard loopmath". The agent runs a dry run, asks you one question (which model labels your sessions, with the cost of each option), onboards, fits, and opens the results page.
75
77
 
76
- `skill install` writes six skills, one per job. Each skill's folder also gets its own copy of the shared `reference.md`, with every command and JSON field the skills use, so each skill works on its own. An agent loads only the skill for the job at hand, and each phrase belongs to one skill:
78
+ `skill install` writes one skill, `loopmath`. It tells the agent to check the store with `loopmath status --json`, pick the job, and read that job's guide from the CLI, so the instructions always match the loopmath you have installed. Mention loopmath in what you ask, and the skill picks the guide:
77
79
 
78
- | Skill | Ask your agent |
80
+ | Guide | Ask your agent |
79
81
  |---|---|
80
- | `loopmath` | loopmath named without a job: "use loopmath", "what can loopmath do", "what next with loopmath" |
81
- | `loopmath-onboard` | "onboard", "set up", "start" or "try" loopmath; or loopmath has no runs yet |
82
- | `loopmath-import-runs` | OCP files (`*.ocp.json`), a folder of finished runs, or the output of an orchestrator or experiment harness that writes OCP |
83
- | `loopmath-update-fit` | "update", "refresh" or "rerun" the fit; what loopmath learned; runs imported or recorded without a refit |
84
- | `loopmath-plan-task` | "plan a task with loopmath"; which workflow, model or effort to use; a coding task of more than a few minutes in a repo that uses loopmath |
85
- | `loopmath-record-run` | a task planned with loopmath is done; "record", "log" or "save" a run; both runs of a pair need a blind judge |
82
+ | `loopmath guide onboard` | "onboard", "set up", "start" or "try" loopmath; or loopmath has no runs yet |
83
+ | `loopmath guide import` | OCP files (`*.ocp.json`), a folder of finished runs, or the output of an orchestrator or experiment harness that writes OCP |
84
+ | `loopmath guide plan` | "plan a task with loopmath"; which workflow, model or effort to use |
85
+ | `loopmath guide record` | a task planned with loopmath is done; "record", "log" or "save" a run; both runs of a pair need a blind judge |
86
+ | `loopmath guide fit` | "update", "refresh" or "rerun" the fit; what loopmath learned; runs imported or recorded without a refit |
87
+
88
+ `loopmath guide` alone lists them, and `loopmath guide reference` prints the command reference they share, with every command, flag and JSON field they use. A guide is read only: it needs no store and writes nothing.
86
89
 
87
- By default the skills go to every agent whose home exists: Claude Code (`~/.claude/skills/<name>/`, or under `$CLAUDE_CONFIG_DIR`) and Codex (`~/.codex/skills/<name>/`, or under `$CODEX_HOME`). `--target claude-code`, `codex` or `both` picks them yourself. If your Codex has no skills folder, the files go to `~/.codex/loopmath/<name>/` instead, with a short marked block in `~/.codex/AGENTS.md` that lists them. Add `--scope project` to install into the current repository (or `--dir PATH` for another one).
90
+ By default the skill goes to every agent whose home exists: Claude Code (`~/.claude/skills/loopmath/`, or under `$CLAUDE_CONFIG_DIR`) and Codex (`~/.codex/skills/loopmath/`, or under `$CODEX_HOME`). `--target claude-code`, `codex` or `both` picks them yourself. If your Codex has no skills folder, the skill goes to `~/.codex/loopmath/loopmath/` instead, with a short marked block in `~/.codex/AGENTS.md` that points to it. Add `--scope project` to install into the current repository (or `--dir PATH` for another one).
88
91
 
89
- It is safe to run again. It records the files it wrote, with their hashes, in `.loopmath-skills.json` beside the skill folders, and it replaces or removes only those. A skill file you changed is kept, with a note. Any other file where a skill goes stops the install before it writes anything, and names the path to move aside. The single skill that 0.1 installed is replaced when it is as 0.1 wrote it. `loopmath skill show NAME` prints one skill, and `loopmath skill uninstall` removes the files it wrote. `loopmath doctor` shows what loopmath can see: logs, agents, store, fit, prices and skills. To read the logs it builds the log parse cache in the store and keeps it, so the next read is fast, and it says how big the cache is; it changes nothing else in the store.
92
+ It is safe to run again. It records the files it wrote, with their hashes, in `.loopmath-skills.json` beside the skill folder, and it replaces or removes only those. A skill file you changed is kept, with a note. Any other file where the skill goes stops the install before it writes anything, and names the path to move aside. The single skill that 0.1 installed is replaced when it is as 0.1 wrote it. Upgrading from 0.2 or 0.3 removes the five job skills those versions installed and the `reference.md` copy in each skill folder, where loopmath wrote them and they are unchanged; a file you changed is kept, with a note. `--dry-run` lists every file install would write or remove. `loopmath skill show` prints the skill, and `loopmath skill uninstall` removes the files it wrote. `loopmath doctor` shows what loopmath can see: logs, agents, store, fit, prices and the skill, and names any job skill an older loopmath left behind. To read the logs it builds the log parse cache in the store and keeps it, so the next read is fast, and it says how big the cache is; it changes nothing else in the store.
90
93
 
91
94
  To onboard by hand instead:
92
95
 
@@ -112,7 +115,7 @@ The prior ships inside the package: our runs, in the shared form (no titles, pat
112
115
 
113
116
  ## The loop
114
117
 
115
- The plan and record skills take an agent through the loop in three commands:
118
+ The plan and record guides take an agent through the loop in three commands:
116
119
 
117
120
  ```sh
118
121
  loopmath recommend --type bug_fix --repo acme/api --title "Fix the parser crash" --json --brief --html
@@ -176,11 +179,11 @@ The workflow builder is the one page that is served: `loopmath builder --type bu
176
179
  | Plan | `task-types`, `workflows` (list, show, validate, diff), `recommend`, `builder` |
177
180
  | Record | `run` (start, record, import, attempt, artifact, finish, sessions), `outcome`, `budget`, `config` (get, set, unset) |
178
181
  | Learn | `fit`, `onboard`, `share`, `prior` |
179
- | View | `runs`, `posterior`, `status`, `report`, `doctor`, `verify-receipts` |
182
+ | View | `runs`, `posterior`, `status`, `report`, `doctor`, `verify-receipts`, `update` |
180
183
  | Harness testing | `check-change`, `gate` (experimental), `screen` (experimental) |
181
184
  | Task subsets | `tasks pick` (experimental) |
182
185
  | Evolution loops | `adapt openevolve` (experimental), `adapt shinkaevolve` (experimental), `recommend --role` (experimental) |
183
- | Skill | `skill` (install, uninstall, show) |
186
+ | Skill | `skill` (install, uninstall, show), `guide` |
184
187
  | OCP | `ocp` (validate, migrate) |
185
188
  | Log analysis | `analyze`, `graph`, `adapt`, `prices`, `validate-prices` |
186
189
 
@@ -307,7 +310,7 @@ The store is `~/.loopmath`. Set `LOOPMATH_HOME` or pass `--home` to put it elsew
307
310
  - signals, receipts and fits;
308
311
  - your config and your workflows;
309
312
  - the views you write;
310
- - the caches: the log parse cache, and the folder where `fit --full` compiles its model (when that compile fails, it also leaves the C code it tried in a `pytensor_compilation_error_*` file in the system temp folder and prints where).
313
+ - the caches: the log parse cache, the daily update check's last answer (`update.json`), and the folder where `fit --full` compiles its model (when that compile fails, it also leaves the C code it tried in a `pytensor_compilation_error_*` file in the system temp folder and prints where).
311
314
 
312
315
  Do not edit the store by hand; use the commands. `loopmath verify-receipts` checks the receipts in your store; give a file to verify a research receipt ledger instead.
313
316
 
@@ -4,7 +4,7 @@ loopmath helps an orchestrator agent choose how to run a coding task, and learns
4
4
 
5
5
  Before a task, it predicts how the user's usual agent workflow will do: the chance of an accepted result, and the cost in tokens and dollars. It shows the other workflows on a success-cost curve and offers two exploration picks. Each pick comes with its gain on every future similar run, its one-time price, and the number of runs it takes to pay back. After the task, the orchestrator records what ran and what came of it. loopmath fills in the token costs from the Claude Code and Codex logs, writes a receipt of predicted against actual, and refits.
6
6
 
7
- loopmath never starts or coordinates agents. Claude Code, Codex, herdr or your own tool does that; loopmath is a CLI they call. Everything runs locally. loopmath connects to no other machine: `loopmath builder` serves its page on 127.0.0.1 only, and no other command opens a network connection. The logs are opened read only, and loopmath writes only to its store (`~/.loopmath`, which holds its caches too), to the skill folders `skill install` names, and to paths you name.
7
+ loopmath never starts or coordinates agents. Claude Code, Codex, herdr or your own tool does that; loopmath is a CLI they call. Everything runs locally. loopmath connects to no other machine but pypi.org: `loopmath builder` serves its page on 127.0.0.1 only, and no other command opens a network connection, except the daily update check. Once a day loopmath asks pypi.org whether a newer loopmath exists; it sends only its version and uploads nothing. `LOOPMATH_NO_UPDATE_CHECK=1` turns the check off, and it is off when `CI` is set. The logs are opened read only, and loopmath writes only to its store (`~/.loopmath`, which holds its caches too), to the skill folders `skill install` names, and to paths you name.
8
8
 
9
9
  The package installs two commands: `loopmath` and its short alias `loop`. If another `loop` is already on your PATH, use `loopmath`.
10
10
 
@@ -26,6 +26,8 @@ python3 -m venv .venv
26
26
 
27
27
  Installing needs the network: pip, pipx and uv download the build backend (setuptools 77 or newer) and the dependencies (jsonschema, numpy, pandas, scipy), from a clone too, so `pip install --no-index -e .` fails offline.
28
28
 
29
+ To upgrade, `loopmath update` says whether a newer loopmath is out and prints the command for how you installed it (`uv tool upgrade loopmath`, `pipx upgrade loopmath`, or pip); run it, then `loopmath skill install`. loopmath never upgrades itself.
30
+
29
31
  ## First run
30
32
 
31
33
  ```sh
@@ -35,20 +37,21 @@ loopmath skill install # teach your coding agents how to use loopmat
35
37
 
36
38
  Then ask your agent to onboard, for example "onboard loopmath". The agent runs a dry run, asks you one question (which model labels your sessions, with the cost of each option), onboards, fits, and opens the results page.
37
39
 
38
- `skill install` writes six skills, one per job. Each skill's folder also gets its own copy of the shared `reference.md`, with every command and JSON field the skills use, so each skill works on its own. An agent loads only the skill for the job at hand, and each phrase belongs to one skill:
40
+ `skill install` writes one skill, `loopmath`. It tells the agent to check the store with `loopmath status --json`, pick the job, and read that job's guide from the CLI, so the instructions always match the loopmath you have installed. Mention loopmath in what you ask, and the skill picks the guide:
39
41
 
40
- | Skill | Ask your agent |
42
+ | Guide | Ask your agent |
41
43
  |---|---|
42
- | `loopmath` | loopmath named without a job: "use loopmath", "what can loopmath do", "what next with loopmath" |
43
- | `loopmath-onboard` | "onboard", "set up", "start" or "try" loopmath; or loopmath has no runs yet |
44
- | `loopmath-import-runs` | OCP files (`*.ocp.json`), a folder of finished runs, or the output of an orchestrator or experiment harness that writes OCP |
45
- | `loopmath-update-fit` | "update", "refresh" or "rerun" the fit; what loopmath learned; runs imported or recorded without a refit |
46
- | `loopmath-plan-task` | "plan a task with loopmath"; which workflow, model or effort to use; a coding task of more than a few minutes in a repo that uses loopmath |
47
- | `loopmath-record-run` | a task planned with loopmath is done; "record", "log" or "save" a run; both runs of a pair need a blind judge |
44
+ | `loopmath guide onboard` | "onboard", "set up", "start" or "try" loopmath; or loopmath has no runs yet |
45
+ | `loopmath guide import` | OCP files (`*.ocp.json`), a folder of finished runs, or the output of an orchestrator or experiment harness that writes OCP |
46
+ | `loopmath guide plan` | "plan a task with loopmath"; which workflow, model or effort to use |
47
+ | `loopmath guide record` | a task planned with loopmath is done; "record", "log" or "save" a run; both runs of a pair need a blind judge |
48
+ | `loopmath guide fit` | "update", "refresh" or "rerun" the fit; what loopmath learned; runs imported or recorded without a refit |
49
+
50
+ `loopmath guide` alone lists them, and `loopmath guide reference` prints the command reference they share, with every command, flag and JSON field they use. A guide is read only: it needs no store and writes nothing.
48
51
 
49
- By default the skills go to every agent whose home exists: Claude Code (`~/.claude/skills/<name>/`, or under `$CLAUDE_CONFIG_DIR`) and Codex (`~/.codex/skills/<name>/`, or under `$CODEX_HOME`). `--target claude-code`, `codex` or `both` picks them yourself. If your Codex has no skills folder, the files go to `~/.codex/loopmath/<name>/` instead, with a short marked block in `~/.codex/AGENTS.md` that lists them. Add `--scope project` to install into the current repository (or `--dir PATH` for another one).
52
+ By default the skill goes to every agent whose home exists: Claude Code (`~/.claude/skills/loopmath/`, or under `$CLAUDE_CONFIG_DIR`) and Codex (`~/.codex/skills/loopmath/`, or under `$CODEX_HOME`). `--target claude-code`, `codex` or `both` picks them yourself. If your Codex has no skills folder, the skill goes to `~/.codex/loopmath/loopmath/` instead, with a short marked block in `~/.codex/AGENTS.md` that points to it. Add `--scope project` to install into the current repository (or `--dir PATH` for another one).
50
53
 
51
- It is safe to run again. It records the files it wrote, with their hashes, in `.loopmath-skills.json` beside the skill folders, and it replaces or removes only those. A skill file you changed is kept, with a note. Any other file where a skill goes stops the install before it writes anything, and names the path to move aside. The single skill that 0.1 installed is replaced when it is as 0.1 wrote it. `loopmath skill show NAME` prints one skill, and `loopmath skill uninstall` removes the files it wrote. `loopmath doctor` shows what loopmath can see: logs, agents, store, fit, prices and skills. To read the logs it builds the log parse cache in the store and keeps it, so the next read is fast, and it says how big the cache is; it changes nothing else in the store.
54
+ It is safe to run again. It records the files it wrote, with their hashes, in `.loopmath-skills.json` beside the skill folder, and it replaces or removes only those. A skill file you changed is kept, with a note. Any other file where the skill goes stops the install before it writes anything, and names the path to move aside. The single skill that 0.1 installed is replaced when it is as 0.1 wrote it. Upgrading from 0.2 or 0.3 removes the five job skills those versions installed and the `reference.md` copy in each skill folder, where loopmath wrote them and they are unchanged; a file you changed is kept, with a note. `--dry-run` lists every file install would write or remove. `loopmath skill show` prints the skill, and `loopmath skill uninstall` removes the files it wrote. `loopmath doctor` shows what loopmath can see: logs, agents, store, fit, prices and the skill, and names any job skill an older loopmath left behind. To read the logs it builds the log parse cache in the store and keeps it, so the next read is fast, and it says how big the cache is; it changes nothing else in the store.
52
55
 
53
56
  To onboard by hand instead:
54
57
 
@@ -74,7 +77,7 @@ The prior ships inside the package: our runs, in the shared form (no titles, pat
74
77
 
75
78
  ## The loop
76
79
 
77
- The plan and record skills take an agent through the loop in three commands:
80
+ The plan and record guides take an agent through the loop in three commands:
78
81
 
79
82
  ```sh
80
83
  loopmath recommend --type bug_fix --repo acme/api --title "Fix the parser crash" --json --brief --html
@@ -138,11 +141,11 @@ The workflow builder is the one page that is served: `loopmath builder --type bu
138
141
  | Plan | `task-types`, `workflows` (list, show, validate, diff), `recommend`, `builder` |
139
142
  | Record | `run` (start, record, import, attempt, artifact, finish, sessions), `outcome`, `budget`, `config` (get, set, unset) |
140
143
  | Learn | `fit`, `onboard`, `share`, `prior` |
141
- | View | `runs`, `posterior`, `status`, `report`, `doctor`, `verify-receipts` |
144
+ | View | `runs`, `posterior`, `status`, `report`, `doctor`, `verify-receipts`, `update` |
142
145
  | Harness testing | `check-change`, `gate` (experimental), `screen` (experimental) |
143
146
  | Task subsets | `tasks pick` (experimental) |
144
147
  | Evolution loops | `adapt openevolve` (experimental), `adapt shinkaevolve` (experimental), `recommend --role` (experimental) |
145
- | Skill | `skill` (install, uninstall, show) |
148
+ | Skill | `skill` (install, uninstall, show), `guide` |
146
149
  | OCP | `ocp` (validate, migrate) |
147
150
  | Log analysis | `analyze`, `graph`, `adapt`, `prices`, `validate-prices` |
148
151
 
@@ -269,7 +272,7 @@ The store is `~/.loopmath`. Set `LOOPMATH_HOME` or pass `--home` to put it elsew
269
272
  - signals, receipts and fits;
270
273
  - your config and your workflows;
271
274
  - the views you write;
272
- - the caches: the log parse cache, and the folder where `fit --full` compiles its model (when that compile fails, it also leaves the C code it tried in a `pytensor_compilation_error_*` file in the system temp folder and prints where).
275
+ - the caches: the log parse cache, the daily update check's last answer (`update.json`), and the folder where `fit --full` compiles its model (when that compile fails, it also leaves the C code it tried in a `pytensor_compilation_error_*` file in the system temp folder and prints where).
273
276
 
274
277
  Do not edit the store by hand; use the commands. `loopmath verify-receipts` checks the receipts in your store; give a file to verify a research receipt ledger instead.
275
278
 
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
4
4
 
5
5
  [project]
6
6
  name = "loopmath"
7
- version = "0.3.3"
7
+ version = "0.3.5"
8
8
  description = "Plan agent workflows, record how they went, and learn from your own Claude Code and Codex logs"
9
9
  readme = "README.md"
10
10
  license = "MIT"
@@ -88,7 +88,7 @@ where = ["src"]
88
88
  loopmath = [
89
89
  "prices.toml",
90
90
  "workflows/catalog/*.toml",
91
- "skill/skills/*.md",
91
+ "skill/guides/*.md",
92
92
  "skill/skills/*/SKILL.md",
93
93
  "priors/*.toml",
94
94
  "priors/bundle/*",
@@ -5,4 +5,4 @@ The command line (`loopmath`, or its alias `loop`) is the interface; see the
5
5
  README and `loopmath --help`.
6
6
  """
7
7
 
8
- __version__ = "0.3.3"
8
+ __version__ = "0.3.5"
@@ -14,9 +14,11 @@ Verbs:
14
14
  plan: task-types workflows recommend builder plan
15
15
  record: run start|record|import|attempt|artifact|finish|sessions outcome budget config
16
16
  learn: fit onboard share
17
- view: runs posterior status report doctor skill ocp
17
+ view: runs posterior status report doctor guide skill update ocp
18
18
 
19
- No verb in this CLI connects to another machine; builder serves on 127.0.0.1 only.
19
+ No verb in this CLI connects to another machine, except the daily update check (update.py): once a
20
+ day, after any command, loopmath asks pypi.org for its latest version, sending only its own
21
+ (LOOPMATH_NO_UPDATE_CHECK=1 or CI turns it off). builder serves on 127.0.0.1 only.
20
22
  Verbs write only to the store ($LOOPMATH_HOME, default ~/.loopmath, which also
21
23
  holds the parse cache and the fit --full compile folder), to the skill folders
22
24
  skill install names, and to paths the user names.
@@ -550,9 +552,23 @@ def _add_verify_receipts(sub) -> None:
550
552
 
551
553
 
552
554
  def main(argv: list[str] | None = None) -> int:
553
- from .cli_errors import JsonErrorParser, run_guarded, wants_json
555
+ """Run one command, then the daily update check (update.py), which never changes the exit code or stdout.
556
+ A command that ends in an exception other than SystemExit, or in Ctrl-C, skips the check."""
557
+ from .update import after_command
554
558
 
555
559
  raw_args = sys.argv[1:] if argv is None else argv
560
+ try:
561
+ code = _main(raw_args)
562
+ except SystemExit: # --help, --version and usage errors too
563
+ after_command(raw_args)
564
+ raise
565
+ after_command(raw_args)
566
+ return code
567
+
568
+
569
+ def _main(raw_args: list[str]) -> int:
570
+ from .cli_errors import JsonErrorParser, run_guarded, wants_json
571
+
556
572
  JsonErrorParser.json_mode = wants_json(raw_args) # under --json a usage error is a JSON object on stdout too
557
573
  JsonErrorParser.argv = list(raw_args)
558
574
  parser = JsonErrorParser(
@@ -633,9 +633,23 @@ def _add_viewing(sub) -> None:
633
633
  _common(p)
634
634
  p.set_defaults(func=_lazy("loopmath.skill.commands:doctor"))
635
635
 
636
- p = sub.add_parser("skill", help="install, remove or print the orchestrator skills")
636
+ p = sub.add_parser("guide", help="print the instructions for one job (onboard, import, plan, record, fit) or "
637
+ "the command reference, for an agent to follow; read only")
638
+ p.add_argument("topic", nargs="?", default=None, metavar="TOPIC",
639
+ help="onboard, import, plan, record, fit or reference (default: list them, with when to use each)")
640
+ p.set_defaults(func=_lazy("loopmath.skill.guide:command"))
641
+
642
+ p = sub.add_parser("update", help="is a newer loopmath out? the installed and latest versions and the upgrade "
643
+ "command; never upgrades")
644
+ p.add_argument("--refresh", action="store_true", help="ask pypi.org now, instead of reusing the last answer "
645
+ "(kept for 24 h, a failed check for 6 h)")
646
+ _common(p)
647
+ p.set_defaults(func=_lazy("loopmath.update:command"))
648
+
649
+ p = sub.add_parser("skill", help="install, remove or print the loopmath skill")
637
650
  ss = p.add_subparsers(dest="skill_command", required=True)
638
- for verb, helptext in (("install", "write the six skills for Claude Code, Codex or both"),
651
+ for verb, helptext in (("install", "write the loopmath skill for Claude Code, Codex or both, and remove the job "
652
+ "skills 0.3.3 and older installed"),
639
653
  ("uninstall", "remove what install wrote")):
640
654
  q = ss.add_parser(verb, help=helptext)
641
655
  q.add_argument("--target", default="auto", choices=["auto", "claude-code", "codex", "both"],
@@ -645,14 +659,12 @@ def _add_viewing(sub) -> None:
645
659
  q.add_argument("--dir", default=None, metavar="PATH", help="the project folder for --scope project "
646
660
  "(default: this folder)")
647
661
  if verb == "install":
648
- q.add_argument("--dry-run", action="store_true", help="print every file install would write or "
649
- "change, and write nothing")
662
+ q.add_argument("--dry-run", action="store_true", help="print every file install would write, change "
663
+ "or remove, and write nothing")
650
664
  _common(q)
651
665
  q.set_defaults(func=_lazy(f"loopmath.skill.commands:{verb}"))
652
- q = ss.add_parser("show", help="print one skill (default loopmath, the start-here skill) or the shared reference")
653
- q.add_argument("name", nargs="?", default="loopmath", metavar="NAME",
654
- help="loopmath, loopmath-onboard, loopmath-import-runs, loopmath-update-fit, loopmath-plan-task, "
655
- "loopmath-record-run, or reference")
666
+ q = ss.add_parser("show", help="print the loopmath skill (the job guides are `loopmath guide`)")
667
+ q.add_argument("name", nargs="?", default="loopmath", metavar="NAME", help="loopmath, the one skill")
656
668
  q.set_defaults(func=_lazy("loopmath.skill.commands:show"))
657
669
 
658
670
  p = sub.add_parser("ocp", help="validate or migrate OCP documents")
@@ -127,7 +127,7 @@ def onboard(args: argparse.Namespace) -> int:
127
127
  print(f"onboard: reading Claude Code and Codex history, {_window_words(since)}"
128
128
  + ("" if H.parse_cache().is_file() else "; a large history can take several minutes on the first read"),
129
129
  file=sys.stderr, flush=True)
130
- # Then, also when captured, what the answers start from, once (the skill shows this line to the user).
130
+ # Then, also when captured, what the answers start from, once (the onboard guide shows this line to the user).
131
131
  prior = starting_prior()
132
132
  print(prior["line"], file=sys.stderr, flush=True)
133
133
  hist = deps.load_history(since_days, logs=deps.logs, progress=_progress, stage=_stage)
@@ -143,15 +143,20 @@ source = "OpenCode 1.17.11 model catalog, opencode free route"
143
143
 
144
144
  # ---------------------------------------------------------------------------
145
145
  # Rows checked on the provider pages on 2026-09-23; gpt-6.1-sol added and
146
- # gpt-6-luna checked again on 2026-10-12 (no newer luna is listed). as_of
147
- # stays the copied table's date: a row with no "checked" date takes it as
148
- # the date of its tariff, and those rows were not read again.
146
+ # gpt-6-luna checked again on 2026-10-12 (no newer luna is listed);
147
+ # claude-sonnet-5-5 and claude-haiku-5-5 confirmed and gpt-6.1-terra added
148
+ # on 2026-11-03. as_of stays the copied table's date: a row with no
149
+ # "checked" date takes it as the date of its tariff, and those rows were
150
+ # not read again.
149
151
  # Each source names its page and the date it was read; the tariff record
150
152
  # written with every priced attempt takes its date from "checked YYYY-MM-DD".
151
153
  #
152
154
  # Anthropic (https://platform.claude.com/docs/en/about-claude/pricing): cache_write is the 1-hour write rate (2x input), as
153
155
  # for the rows above. Claude Code writes the 1-hour cache: 64.2M of 64.2M
154
156
  # cache-write tokens in 200 recent sessions on the build machine were 1h.
157
+ # Claude Haiku 5.5 is priced by prompt length: its row is the tier for
158
+ # prompts up to 100,000 input tokens, a longer prompt bills 5x those rates
159
+ # and is priced at the same rate here, so a Haiku 5.5 figure is a floor.
155
160
  # OpenAI (https://developers.openai.com/api/docs/pricing and each model's
156
161
  # page): GPT-6 bills cache writes (1.25x input), but Codex reported 0
157
162
  # cache-write tokens in all 91 recent GPT-6 sessions on the build machine,
@@ -168,6 +173,20 @@ cache_write = 8.00
168
173
  output = 20.00
169
174
  source = "official list, https://platform.claude.com/docs/en/about-claude/pricing, checked 2026-09-23: Claude Opus 5.5 at $4 input, $8 1h cache write, $0.20 cache read (0.05x input), $20 output; cache_write is the 1h rate"
170
175
 
176
+ ["claude-sonnet-5-5"]
177
+ input = 2.00
178
+ cache_read = 0.10
179
+ cache_write = 4.00
180
+ output = 10.00
181
+ source = "official list, https://platform.claude.com/docs/en/about-claude/pricing, checked 2026-11-03: Claude Sonnet 5.5 at $2 input, $4 1h cache write, $0.10 cache read (0.05x input), $10 output; cache_write is the 1h rate"
182
+
183
+ ["claude-haiku-5-5"]
184
+ input = 0.10
185
+ cache_read = 0.01
186
+ cache_write = 0.20
187
+ output = 0.50
188
+ source = "official list, https://platform.claude.com/docs/en/about-claude/pricing, checked 2026-11-03: Claude Haiku 5.5 at $0.10 input, $0.20 1h cache write, $0.01 cache read, $0.50 output for prompts up to 100,000 tokens; prompts over 100,000 tokens bill $0.50 input, $1 1h cache write, $0.05 cache read, $2.50 output and are priced at this rate (a floor); cache_write is the 1h rate"
189
+
171
190
  ["gpt-6-astra"]
172
191
  input = 10.00
173
192
  cache_read = 1.00
@@ -189,6 +208,13 @@ cache_write = 2.50
189
208
  output = 10.00
190
209
  source = "official list, https://developers.openai.com/api/docs/models/gpt-6.1-sol, checked 2026-10-12: standard tier, $2 input, $0.10 cached input, $2.50 cache write, $10 output; prompts over 272K input tokens bill more and are priced at this rate (a floor)"
191
210
 
211
+ ["gpt-6.1-terra"]
212
+ input = 1.00
213
+ cache_read = 0.05
214
+ cache_write = 1.25
215
+ output = 5.00
216
+ source = "official list, https://developers.openai.com/api/docs/models/gpt-6.1-terra, checked 2026-11-03: standard tier, $1 input, $0.05 cached input, $1.25 cache write (30-minute retention; 12-hour retention is $2.50), $5 output; prompts over 272K input tokens bill more and are priced at this rate (a floor)"
217
+
192
218
  ["gpt-6-luna"]
193
219
  input = 0.10
194
220
  cache_read = 0.01
@@ -1,4 +1,5 @@
1
1
  """Handlers for `loopmath doctor` and `loopmath skill install|uninstall|show`.
2
+ (`loopmath guide` is guide.py's, which imports nothing else.)
2
3
  """
3
4
 
4
5
  from __future__ import annotations
@@ -36,16 +37,9 @@ _WOULD = {"created": "to write", "updated": "to update", "replaced": "to replace
36
37
  "removed": "to remove", "unchanged": "unchanged", "kept": "kept"}
37
38
 
38
39
 
39
- def _count(entries: list[dict], dry_run: bool = False) -> str:
40
- counts: dict[str, int] = {}
41
- for e in entries:
42
- counts[e["action"]] = counts.get(e["action"], 0) + 1
43
- return ", ".join(f"{n} {_WOULD[action] if dry_run else action}" for action, n in counts.items())
44
-
45
-
46
40
  def _results(verb: str, results: list[dict], as_json: bool) -> None:
47
41
  if as_json:
48
- emit_json("loopmath.skill/2", {"verb": verb, "results": results})
42
+ emit_json("loopmath.skill/3", {"verb": verb, "results": results})
49
43
  return
50
44
  dry = any(r.get("dry_run") for r in results)
51
45
  if dry:
@@ -53,19 +47,20 @@ def _results(verb: str, results: list[dict], as_json: bool) -> None:
53
47
  for r in results:
54
48
  form = " (AGENTS.md block)" if r["method"] == "agents_block" else ""
55
49
  head = f"{r['target']} {r['scope']}{form}"
56
- actions = {e["action"] for e in r["skills"] + r["reference"]}
57
- if verb == "uninstall" and actions <= {"absent"}:
50
+ [skill] = r["skills"]
51
+ if verb == "uninstall" and skill["action"] == "absent" and not r["removed"]:
58
52
  print(f"{head}: not installed in {r['root']}, nothing to remove")
59
- elif verb == "uninstall" and actions <= {"removed", "absent"}:
60
- n = sum(e["action"] == "removed" for e in r["skills"])
61
- print(f"{head}: removed {n} skill{'s' if n != 1 else ''} and the reference.md beside each from {r['root']}")
62
- else:
63
- print(f"{head}: {len(r['skills'])} skills in {r['root']}: {_count(r['skills'], dry)}")
64
- if dry:
65
- print(f" and in each skill folder a copy of the shared reference.md ({_count(r['reference'], dry)}), "
66
- f"and {r['manifest']}, the list of the files install wrote")
53
+ continue
54
+ print(f"{head}:")
55
+ if skill["action"] != "absent":
56
+ print(f" {_WOULD[skill['action']] if dry else skill['action']} {skill['path']}, the loopmath skill")
57
+ action = r.get("manifest_action")
58
+ if dry and action:
59
+ print(f" {_WOULD[action]} {r['manifest']}, the list of the files install wrote")
67
60
  for e in r["removed"]:
68
61
  print(f" {'to remove' if dry else 'removed'} {e['path']}")
62
+ if any(e["path"].endswith(f"/{name}/SKILL.md") for e in r["removed"] for name in _install.RETIRED_SKILLS):
63
+ print(" (0.3.4 installs one skill; the job skills' text comes from `loopmath guide` now)")
69
64
  block, where = r.get("block"), r.get("block_file") or r.get("agents_md")
70
65
  if block == "removed":
71
66
  print(f" {'to remove' if dry else 'removed'} the loopmath block from {where}")
@@ -104,5 +99,8 @@ def show(args: argparse.Namespace) -> int:
104
99
  try:
105
100
  sys.stdout.write(_install.skill_text(name))
106
101
  except KeyError:
107
- return fail(f"no skill {name!r}; one of {', '.join(_install.SKILLS)}, or reference")
102
+ topic = "reference" if name == "reference" else _install.RETIRED_SKILLS.get(name)
103
+ if topic:
104
+ return fail(f"0.3.4 has one skill, loopmath; {name} is `loopmath guide {topic}` now")
105
+ return fail(f"no skill {name!r}; the one skill is loopmath, and `loopmath guide` lists the job guides")
108
106
  return EXIT_OK
@@ -6,9 +6,11 @@ Six checks, each `ok`, `warn` or `fail`:
6
6
  store the store root exists and is writable
7
7
  fit `fits/latest` present, and its age
8
8
  prices every model seen in the scan window has a price
9
- skill where the skill set is installed (and which Codex form), and whether it is whole and current
9
+ skill where the loopmath skill is installed (and which Codex form), whether it is the one this
10
+ loopmath writes, and the job skills 0.3.3 and older installed that are still on disk
10
11
 
11
- Nothing here opens a network connection. Log files are read through the
12
+ Nothing here opens a network connection (the daily update check runs after the command, from
13
+ cli.main). Log files are read through the
12
14
  parse cache, which doctor keeps up to date in the store's cache folder (the
13
15
  report's `cache` says where and how large); it writes nothing else there, and
14
16
  creates the store folder only when it is missing. Codex auth files are never
@@ -28,6 +30,7 @@ from collections import Counter
28
30
  from pathlib import Path
29
31
  from typing import Callable, Mapping
30
32
 
33
+ from .. import __version__
31
34
  from ..belief.design import other_design
32
35
  from . import install as _install
33
36
 
@@ -189,28 +192,52 @@ def _install_cmd(place: dict) -> str:
189
192
  return f"loopmath skill install --target {place['target']} --scope {place['scope']}"
190
193
 
191
194
 
195
+ def _folders(paths: list[str]) -> str:
196
+ """The retired files, by the skill folder each is in (`loopmath/reference.md` by its name)."""
197
+ names = []
198
+ for path in paths:
199
+ folder, file = Path(path).parent.name, Path(path).name
200
+ name = f"{folder}/{file}" if folder == "loopmath" else folder
201
+ if name not in names:
202
+ names.append(name)
203
+ return ", ".join(names)
204
+
205
+
192
206
  def check_skill(cwd: Path | None, env: Mapping[str, str] | None) -> dict:
193
- """`ok` when the full set is installed and current for some target, with nothing old or partial
194
- left anywhere; else a warning with the exact install command."""
207
+ """`ok` when the skill is installed and current for some target, with nothing old, partial or
208
+ retired left anywhere; else a warning with the exact install command. Retired files loopmath still
209
+ owns go with that command; the others were kept, and the warning says why."""
195
210
  places = _install.status(cwd=cwd, env=env)
196
- n = len(_install.SKILLS)
197
211
  seen = [p for p in places if p["any"]]
198
212
  if not seen:
199
213
  return _check("skill", "warn", "not installed: run `loopmath skill install`", places=places)
200
214
  problems = []
201
215
  for p in seen:
202
216
  where = f"{p['target']} {p['scope']}"
217
+ retired = p["retired"]
203
218
  if p["old"]:
204
219
  problems.append(f"{where} still has the 0.1 single skill: run `{_install_cmd(p)}`")
205
220
  elif p["missing"]:
206
- problems.append(f"{where} lacks {', '.join(p['missing'])}: run `{_install_cmd(p)}`")
221
+ problems.append(f"{where} lacks the loopmath skill: run `{_install_cmd(p)}`")
207
222
  elif not p["current"]:
208
- problems.append(f"{where} is older than this loopmath: run `{_install_cmd(p)}`")
223
+ problems.append(f"{where} has a loopmath skill that differs from the one loopmath {__version__} writes: "
224
+ f"run `{_install_cmd(p)}`")
225
+ if retired["owned"]:
226
+ problems.append(f"{where} still has the job skills 0.3.3 and older installed "
227
+ f"({_folders(retired['owned'])}): run `{_install_cmd(p)}` to remove them")
228
+ if retired["changed"]:
229
+ problems.append(f"{where} kept {', '.join(retired['changed'])}: changed since loopmath wrote "
230
+ f"{'them' if len(retired['changed']) > 1 else 'it'}; move "
231
+ f"{'them' if len(retired['changed']) > 1 else 'it'} aside when no longer needed")
232
+ if retired["foreign"]:
233
+ problems.append(f"{where} has {', '.join(retired['foreign'])}, where 0.3.3 and older put job skills, "
234
+ "but loopmath did not write "
235
+ f"{'them' if len(retired['foreign']) > 1 else 'it'}: left alone")
209
236
  if problems:
210
237
  return _check("skill", "warn", "; ".join(problems), places=places)
211
238
  parts = [f"{p['target']} {p['scope']} ({'AGENTS.md block' if p['method'] == 'agents_block' else 'skills'})"
212
239
  for p in seen]
213
- return _check("skill", "ok", f"{n} skills: " + ", ".join(parts), places=places)
240
+ return _check("skill", "ok", "the loopmath skill: " + ", ".join(parts), places=places)
214
241
 
215
242
 
216
243
  def run_checks(home: Path, *, env: Mapping[str, str] | None = None, cwd: Path | None = None,
@@ -0,0 +1,52 @@
1
+ """`loopmath guide [TOPIC]`: the instructions for each job, for the agent to read and follow.
2
+
3
+ The one installed skill (`skills/loopmath/SKILL.md`) sends the agent here, so a job's
4
+ instructions always match the installed loopmath. Each topic is a packaged file in `guides/`:
5
+ the five jobs, then the command reference they share. Read only: it needs no store and writes
6
+ nothing, so this module imports nothing from the rest of loopmath.
7
+ """
8
+
9
+ from __future__ import annotations
10
+
11
+ import argparse
12
+ import sys
13
+ from importlib import resources
14
+
15
+ # Each topic, with one line on when to use it, in the order `loopmath guide` lists them.
16
+ TOPICS = {
17
+ "onboard": "set up, start or try loopmath from the user's Claude Code and Codex history, or loopmath has no runs yet",
18
+ "import": "OCP files or a folder of finished runs to bring into loopmath",
19
+ "plan": "a task to do: pick its workflow, models and effort, then start the run",
20
+ "record": "a task is done: record the run and show its receipt, or judge a pair blind",
21
+ "fit": "update the fit, say what loopmath learned, or the fit is not usable",
22
+ "reference": "every command, flag, JSON field and exit code the guides use",
23
+ }
24
+ FOLLOW = "Run `loopmath guide <topic>` and follow it."
25
+ EXIT_UNKNOWN = 2 # the CLI's "not found"
26
+
27
+
28
+ def text(topic: str) -> str:
29
+ """One topic's text, as packaged. KeyError on another name."""
30
+ if topic not in TOPICS:
31
+ raise KeyError(topic)
32
+ return resources.files("loopmath.skill").joinpath("guides", f"{topic}.md").read_text(encoding="utf-8")
33
+
34
+
35
+ def listing() -> str:
36
+ """What `loopmath guide` alone prints: each topic with when to use it, then how to read one."""
37
+ width = max(map(len, TOPICS))
38
+ rows = [f" {name:<{width}} {when}" for name, when in TOPICS.items()]
39
+ return "loopmath guides, one per job, and the command reference:\n\n" + "\n".join(rows) + f"\n\n{FOLLOW}\n"
40
+
41
+
42
+ def command(args: argparse.Namespace) -> int:
43
+ topic = getattr(args, "topic", None)
44
+ if topic is None:
45
+ sys.stdout.write(listing())
46
+ return 0
47
+ try:
48
+ sys.stdout.write(text(topic))
49
+ except KeyError:
50
+ print(f"error: no guide {topic!r}; the topics are {', '.join(TOPICS)}. {FOLLOW}", file=sys.stderr)
51
+ return EXIT_UNKNOWN
52
+ return 0
@@ -1,14 +1,8 @@
1
- ---
2
- name: loopmath-update-fit
3
- description: "Refit loopmath on every run recorded so far and show the results page. Use when the user asks to update, refresh or rerun the loopmath fit, asks what loopmath learned or believes about models, efforts or workflows, or after runs were imported or recorded without a refit."
4
- allowed-tools: Bash(loopmath:*), Bash(open:*), Bash(xdg-open:*)
5
- ---
6
-
7
1
  # Update the loopmath fit
8
2
 
9
3
  Refit on everything in the store, then show the results page. No question.
10
4
 
11
- Use exactly the commands below. Flags and fields are in `reference.md` next to this file. Never run `--help`.
5
+ Use exactly the commands below. Flags and fields are in `loopmath guide reference`. Never run `--help`.
12
6
 
13
7
  ## 1. Fit
14
8
 
@@ -24,7 +18,7 @@ A new fit replaces the one `recommend` and the pages read. Read `fit.id`, `fit.n
24
18
 
25
19
  `shipped_overlap` counts, per shipped source, the user's runs that the shipped prior also holds (the same runs); the fit uses the user's copies. When it has counts, the summary says one line, once: "N of your runs are also in the shipped SOURCE prior (same runs): fits use your copies", with the total for N ("1 of your runs is also in the shipped SOURCE prior (same run): fits use your copy" for one run), and with several sources "the shipped prior (rq1 40, e0 4)" for "the shipped SOURCE prior". Its `dropped` reason, "shipped copy of a stored run", is that same overlap: do not name it again.
26
20
 
27
- If it exits 4 (another fit holds the lock), wait 30 seconds and run it once more. If `fit.n_runs.user` is 0 and the user expected runs, say so and suggest `loopmath-onboard` or `loopmath-import-runs`.
21
+ If it exits 4 (another fit holds the lock), wait 30 seconds and run it once more. If `fit.n_runs.user` is 0 and the user expected runs, say so and suggest onboarding (`loopmath guide onboard`) or bringing existing runs (`loopmath guide import`).
28
22
 
29
23
  ## 2. Show the results page
30
24
 
@@ -1,14 +1,8 @@
1
- ---
2
- name: loopmath-import-runs
3
- description: "Bring agent runs the user already has as OCP files (*.ocp.json) into loopmath in one pass: validate all, import all, fit once, show the results page. Use when the user points at OCP files or a folder of finished runs, or at the output of an orchestrator or experiment harness that writes OCP, and wants loopmath to learn from them."
4
- allowed-tools: Bash(loopmath:*), Bash(open:*), Bash(xdg-open:*)
5
- ---
6
-
7
1
  # Bring existing runs into loopmath
8
2
 
9
3
  The user has finished runs written as OCP v0.3 (or v0.1, v0.2) documents. Validate them all, import them all without refitting after each, fit once, and show the results page. Ask nothing unless the location of the files is unknown.
10
4
 
11
- Use exactly the commands below. Flags and fields are in `reference.md` next to this file. Never run `--help`.
5
+ Use exactly the commands below. Flags and fields are in `loopmath guide reference`. Never run `--help`.
12
6
 
13
7
  ## 1. Find the files
14
8