@team-agent/installer 0.5.127 → 0.6.0

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 (421) hide show
  1. package/README.md +15 -4
  2. package/README.zh.md +1 -1
  3. package/package.json +4 -8
  4. package/skills/team-agent/SKILL.md +41 -26
  5. package/skills/team-agent/command-coverage.json +170 -184
  6. package/skills/team-agent/docs/reference/team-agent-operator.md +66 -32
  7. package/skills/team-agent/references/recovery-runbook.md +1 -1
  8. package/skills/team-agent/references/team-agent-operator.md +66 -32
  9. package/templates/team_state.md +1 -1
  10. package/Cargo.lock +0 -1063
  11. package/Cargo.toml +0 -36
  12. package/crates/team-agent/Cargo.toml +0 -69
  13. package/crates/team-agent/src/app_server_test_support.rs +0 -258
  14. package/crates/team-agent/src/cli/adapters.rs +0 -1298
  15. package/crates/team-agent/src/cli/arguments.rs +0 -401
  16. package/crates/team-agent/src/cli/attach_app_server_leader.rs +0 -98
  17. package/crates/team-agent/src/cli/diagnose.rs +0 -1675
  18. package/crates/team-agent/src/cli/emit.rs +0 -2898
  19. package/crates/team-agent/src/cli/grok_slot.rs +0 -384
  20. package/crates/team-agent/src/cli/helpers.rs +0 -91
  21. package/crates/team-agent/src/cli/intent/runtime.rs +0 -1189
  22. package/crates/team-agent/src/cli/intent/runtime_tests.rs +0 -821
  23. package/crates/team-agent/src/cli/intent/tests.rs +0 -691
  24. package/crates/team-agent/src/cli/intent.rs +0 -664
  25. package/crates/team-agent/src/cli/leader.rs +0 -300
  26. package/crates/team-agent/src/cli/leader_prompt.rs +0 -261
  27. package/crates/team-agent/src/cli/leaders.rs +0 -418
  28. package/crates/team-agent/src/cli/mod.rs +0 -6528
  29. package/crates/team-agent/src/cli/models.rs +0 -449
  30. package/crates/team-agent/src/cli/named_address.rs +0 -1651
  31. package/crates/team-agent/src/cli/profile.rs +0 -342
  32. package/crates/team-agent/src/cli/route.rs +0 -212
  33. package/crates/team-agent/src/cli/send/coordinator.rs +0 -164
  34. package/crates/team-agent/src/cli/send/mailbox.rs +0 -145
  35. package/crates/team-agent/src/cli/send/persist.rs +0 -192
  36. package/crates/team-agent/src/cli/send/presentation.rs +0 -370
  37. package/crates/team-agent/src/cli/send/resolve.rs +0 -573
  38. package/crates/team-agent/src/cli/send.rs +0 -201
  39. package/crates/team-agent/src/cli/shutdown_socket.rs +0 -288
  40. package/crates/team-agent/src/cli/spec.rs +0 -159
  41. package/crates/team-agent/src/cli/status.rs +0 -280
  42. package/crates/team-agent/src/cli/status_port/agents.rs +0 -358
  43. package/crates/team-agent/src/cli/status_port/approvals.rs +0 -79
  44. package/crates/team-agent/src/cli/status_port/brief.rs +0 -2043
  45. package/crates/team-agent/src/cli/status_port/compact.rs +0 -240
  46. package/crates/team-agent/src/cli/status_port/format.rs +0 -146
  47. package/crates/team-agent/src/cli/status_port/inbox.rs +0 -62
  48. package/crates/team-agent/src/cli/status_port/runtime.rs +0 -197
  49. package/crates/team-agent/src/cli/status_port/snapshot.rs +0 -208
  50. package/crates/team-agent/src/cli/status_port/store.rs +0 -416
  51. package/crates/team-agent/src/cli/status_port/tests.rs +0 -90
  52. package/crates/team-agent/src/cli/status_port.rs +0 -92
  53. package/crates/team-agent/src/cli/tests/base.rs +0 -885
  54. package/crates/team-agent/src/cli/tests/divergence.rs +0 -518
  55. package/crates/team-agent/src/cli/tests/lane_c.rs +0 -453
  56. package/crates/team-agent/src/cli/tests/leader_watch.rs +0 -372
  57. package/crates/team-agent/src/cli/tests/main_preserved.rs +0 -430
  58. package/crates/team-agent/src/cli/tests/missing_subcommands.rs +0 -336
  59. package/crates/team-agent/src/cli/tests/mod.rs +0 -137
  60. package/crates/team-agent/src/cli/tests/named_address.rs +0 -760
  61. package/crates/team-agent/src/cli/tests/parser_safety.rs +0 -383
  62. package/crates/team-agent/src/cli/tests/run_delegation.rs +0 -481
  63. package/crates/team-agent/src/cli/tests/shutdown_kill_plan.rs +0 -2179
  64. package/crates/team-agent/src/cli/tests/status_brief.rs +0 -1514
  65. package/crates/team-agent/src/cli/tests/status_send.rs +0 -1091
  66. package/crates/team-agent/src/cli/tests/verb_install_skill.rs +0 -93
  67. package/crates/team-agent/src/cli/tests/verb_profile.rs +0 -392
  68. package/crates/team-agent/src/cli/triage.rs +0 -234
  69. package/crates/team-agent/src/cli/types.rs +0 -591
  70. package/crates/team-agent/src/codex_app_server.rs +0 -546
  71. package/crates/team-agent/src/communication_mode/mod.rs +0 -56
  72. package/crates/team-agent/src/compiler/tests.rs +0 -842
  73. package/crates/team-agent/src/compiler.rs +0 -691
  74. package/crates/team-agent/src/conpty/backend.rs +0 -829
  75. package/crates/team-agent/src/conpty/mod.rs +0 -33
  76. package/crates/team-agent/src/contract_runtime/backend.rs +0 -1610
  77. package/crates/team-agent/src/contract_runtime/commands.rs +0 -671
  78. package/crates/team-agent/src/contract_runtime/config.rs +0 -354
  79. package/crates/team-agent/src/contract_runtime/config_tests.rs +0 -56
  80. package/crates/team-agent/src/contract_runtime/discovery.rs +0 -438
  81. package/crates/team-agent/src/contract_runtime/fork.rs +0 -368
  82. package/crates/team-agent/src/contract_runtime/forward.rs +0 -288
  83. package/crates/team-agent/src/contract_runtime/framework.rs +0 -350
  84. package/crates/team-agent/src/contract_runtime/mcp.rs +0 -399
  85. package/crates/team-agent/src/contract_runtime/mod.rs +0 -17
  86. package/crates/team-agent/src/contract_runtime/prompt.rs +0 -216
  87. package/crates/team-agent/src/contract_runtime/pump.rs +0 -271
  88. package/crates/team-agent/src/contract_runtime/registry.rs +0 -589
  89. package/crates/team-agent/src/contract_runtime/workspace.rs +0 -228
  90. package/crates/team-agent/src/coordinator/backoff.rs +0 -741
  91. package/crates/team-agent/src/coordinator/conpty_shim.rs +0 -805
  92. package/crates/team-agent/src/coordinator/health.rs +0 -1185
  93. package/crates/team-agent/src/coordinator/mod.rs +0 -119
  94. package/crates/team-agent/src/coordinator/orphan.rs +0 -110
  95. package/crates/team-agent/src/coordinator/runtime_detectors.rs +0 -798
  96. package/crates/team-agent/src/coordinator/runtime_observation.rs +0 -84
  97. package/crates/team-agent/src/coordinator/steps/abnormal.rs +0 -2309
  98. package/crates/team-agent/src/coordinator/steps/delivery.rs +0 -15
  99. package/crates/team-agent/src/coordinator/steps/health_sync.rs +0 -15
  100. package/crates/team-agent/src/coordinator/steps/mod.rs +0 -126
  101. package/crates/team-agent/src/coordinator/steps/persist.rs +0 -15
  102. package/crates/team-agent/src/coordinator/steps/runtime_prompts.rs +0 -15
  103. package/crates/team-agent/src/coordinator/steps/session_gate.rs +0 -16
  104. package/crates/team-agent/src/coordinator/tests/a0_lostupdate.rs +0 -158
  105. package/crates/team-agent/src/coordinator/tests/abnormal.rs +0 -312
  106. package/crates/team-agent/src/coordinator/tests/api_error_recovery.rs +0 -338
  107. package/crates/team-agent/src/coordinator/tests/basics.rs +0 -211
  108. package/crates/team-agent/src/coordinator/tests/daemon.rs +0 -514
  109. package/crates/team-agent/src/coordinator/tests/energy.rs +0 -549
  110. package/crates/team-agent/src/coordinator/tests/health_sync.rs +0 -461
  111. package/crates/team-agent/src/coordinator/tests/main_preserved.rs +0 -186
  112. package/crates/team-agent/src/coordinator/tests/mod.rs +0 -295
  113. package/crates/team-agent/src/coordinator/tests/s3_bounded.rs +0 -207
  114. package/crates/team-agent/src/coordinator/tests/spine.rs +0 -308
  115. package/crates/team-agent/src/coordinator/tests/takeover.rs +0 -286
  116. package/crates/team-agent/src/coordinator/tests/tick_core.rs +0 -237
  117. package/crates/team-agent/src/coordinator/tick.rs +0 -3112
  118. package/crates/team-agent/src/coordinator/types.rs +0 -540
  119. package/crates/team-agent/src/db/agent_health_capture.rs +0 -126
  120. package/crates/team-agent/src/db/message_store.rs +0 -1585
  121. package/crates/team-agent/src/db/migration.rs +0 -984
  122. package/crates/team-agent/src/db/mod.rs +0 -26
  123. package/crates/team-agent/src/db/schema.rs +0 -663
  124. package/crates/team-agent/src/diagnose/comms.rs +0 -242
  125. package/crates/team-agent/src/diagnose/mod.rs +0 -44
  126. package/crates/team-agent/src/diagnose/orphans.rs +0 -889
  127. package/crates/team-agent/src/event_log.rs +0 -502
  128. package/crates/team-agent/src/fake_worker.rs +0 -303
  129. package/crates/team-agent/src/kill_audit.rs +0 -106
  130. package/crates/team-agent/src/layout/manager.rs +0 -248
  131. package/crates/team-agent/src/layout/mod.rs +0 -36
  132. package/crates/team-agent/src/layout/overlay.rs +0 -134
  133. package/crates/team-agent/src/layout/placement.rs +0 -52
  134. package/crates/team-agent/src/layout/recovery.rs +0 -106
  135. package/crates/team-agent/src/layout/runtime_sessions.rs +0 -317
  136. package/crates/team-agent/src/layout/sessions.rs +0 -294
  137. package/crates/team-agent/src/layout/tmux_endpoint.rs +0 -165
  138. package/crates/team-agent/src/layout/worker_env.rs +0 -721
  139. package/crates/team-agent/src/layout/worker_window_helpers.rs +0 -37
  140. package/crates/team-agent/src/leader/helpers.rs +0 -178
  141. package/crates/team-agent/src/leader/incident.rs +0 -6
  142. package/crates/team-agent/src/leader/lease.rs +0 -3356
  143. package/crates/team-agent/src/leader/mod.rs +0 -104
  144. package/crates/team-agent/src/leader/owner_bind.rs +0 -400
  145. package/crates/team-agent/src/leader/prompt.rs +0 -905
  146. package/crates/team-agent/src/leader/provider_attribution.rs +0 -247
  147. package/crates/team-agent/src/leader/rediscover/tests.rs +0 -689
  148. package/crates/team-agent/src/leader/rediscover.rs +0 -1192
  149. package/crates/team-agent/src/leader/registry.rs +0 -883
  150. package/crates/team-agent/src/leader/start.rs +0 -4500
  151. package/crates/team-agent/src/leader/takeover.rs +0 -65
  152. package/crates/team-agent/src/leader/tests/basics.rs +0 -337
  153. package/crates/team-agent/src/leader/tests/byte_findings.rs +0 -355
  154. package/crates/team-agent/src/leader/tests/claim_leader_mailbox_flush_contract.rs +0 -544
  155. package/crates/team-agent/src/leader/tests/historical_inventory_upgrade_gate_red.rs +0 -624
  156. package/crates/team-agent/src/leader/tests/identity.rs +0 -465
  157. package/crates/team-agent/src/leader/tests/identity_session_names.rs +0 -42
  158. package/crates/team-agent/src/leader/tests/idle.rs +0 -113
  159. package/crates/team-agent/src/leader/tests/lease_api.rs +0 -520
  160. package/crates/team-agent/src/leader/tests/lease_claim.rs +0 -629
  161. package/crates/team-agent/src/leader/tests/mod.rs +0 -106
  162. package/crates/team-agent/src/leader/tests/prompt_contract.rs +0 -267
  163. package/crates/team-agent/src/leader/tests/rediscover.rs +0 -403
  164. package/crates/team-agent/src/leader/tests/rehomed_env.rs +0 -83
  165. package/crates/team-agent/src/leader/tests/team_in_team_state_scope_red.rs +0 -1273
  166. package/crates/team-agent/src/leader/tests/terminal_and_replay_vs_rerender_invariants_red.rs +0 -417
  167. package/crates/team-agent/src/leader/tests/wake_start_owner.rs +0 -168
  168. package/crates/team-agent/src/leader/types.rs +0 -534
  169. package/crates/team-agent/src/lib.rs +0 -145
  170. package/crates/team-agent/src/lifecycle/helpers.rs +0 -212
  171. package/crates/team-agent/src/lifecycle/launch/add_agent.rs +0 -848
  172. package/crates/team-agent/src/lifecycle/launch/add_agent_state.rs +0 -393
  173. package/crates/team-agent/src/lifecycle/launch/agent_state.rs +0 -332
  174. package/crates/team-agent/src/lifecycle/launch/clone_agent.rs +0 -178
  175. package/crates/team-agent/src/lifecycle/launch/fork_agent.rs +0 -82
  176. package/crates/team-agent/src/lifecycle/launch/fork_entry.rs +0 -63
  177. package/crates/team-agent/src/lifecycle/launch/identity.rs +0 -674
  178. package/crates/team-agent/src/lifecycle/launch/layout.rs +0 -350
  179. package/crates/team-agent/src/lifecycle/launch/leader_context.rs +0 -589
  180. package/crates/team-agent/src/lifecycle/launch/mcp_config.rs +0 -152
  181. package/crates/team-agent/src/lifecycle/launch/ownership.rs +0 -105
  182. package/crates/team-agent/src/lifecycle/launch/pi_mcp.rs +0 -629
  183. package/crates/team-agent/src/lifecycle/launch/pi_mcp_extension.js +0 -226
  184. package/crates/team-agent/src/lifecycle/launch/plan.rs +0 -269
  185. package/crates/team-agent/src/lifecycle/launch/quick_start.rs +0 -5176
  186. package/crates/team-agent/src/lifecycle/launch/quick_start_transport.rs +0 -368
  187. package/crates/team-agent/src/lifecycle/launch/readiness.rs +0 -519
  188. package/crates/team-agent/src/lifecycle/launch/role_source.rs +0 -287
  189. package/crates/team-agent/src/lifecycle/launch/spawn.rs +0 -450
  190. package/crates/team-agent/src/lifecycle/launch/spec_state.rs +0 -801
  191. package/crates/team-agent/src/lifecycle/launch/state_projection.rs +0 -624
  192. package/crates/team-agent/src/lifecycle/launch/worker_env.rs +0 -558
  193. package/crates/team-agent/src/lifecycle/launch.rs +0 -534
  194. package/crates/team-agent/src/lifecycle/lock.rs +0 -350
  195. package/crates/team-agent/src/lifecycle/mod.rs +0 -79
  196. package/crates/team-agent/src/lifecycle/pane_input_lock.rs +0 -239
  197. package/crates/team-agent/src/lifecycle/profile_launch.rs +0 -757
  198. package/crates/team-agent/src/lifecycle/profile_smoke.rs +0 -557
  199. package/crates/team-agent/src/lifecycle/restart/agent.rs +0 -2198
  200. package/crates/team-agent/src/lifecycle/restart/common.rs +0 -2052
  201. package/crates/team-agent/src/lifecycle/restart/orchestrator.rs +0 -65
  202. package/crates/team-agent/src/lifecycle/restart/preflight.rs +0 -113
  203. package/crates/team-agent/src/lifecycle/restart/rebuild.rs +0 -5135
  204. package/crates/team-agent/src/lifecycle/restart/remove.rs +0 -1504
  205. package/crates/team-agent/src/lifecycle/restart/selection.rs +0 -559
  206. package/crates/team-agent/src/lifecycle/restart/team_state.rs +0 -329
  207. package/crates/team-agent/src/lifecycle/restart.rs +0 -223
  208. package/crates/team-agent/src/lifecycle/role_config.rs +0 -1205
  209. package/crates/team-agent/src/lifecycle/tests/acceptance_b_batch_red.rs +0 -1021
  210. package/crates/team-agent/src/lifecycle/tests/agent_ops.rs +0 -1205
  211. package/crates/team-agent/src/lifecycle/tests/b0_legacy_snapshot_nonauthority_contract.rs +0 -265
  212. package/crates/team-agent/src/lifecycle/tests/b0_reader_hideone_audit_contract.rs +0 -490
  213. package/crates/team-agent/src/lifecycle/tests/behavioral_diff_264_red.rs +0 -656
  214. package/crates/team-agent/src/lifecycle/tests/claude_profile_launch_red.rs +0 -135
  215. package/crates/team-agent/src/lifecycle/tests/cli_argv_routing_red.rs +0 -1415
  216. package/crates/team-agent/src/lifecycle/tests/clone_agent_preserves_source_tools.rs +0 -427
  217. package/crates/team-agent/src/lifecycle/tests/clone_fork_copilot_perms_red.rs +0 -396
  218. package/crates/team-agent/src/lifecycle/tests/communication_mode_runtime_contract_red.rs +0 -418
  219. package/crates/team-agent/src/lifecycle/tests/copilot_provider_red.rs +0 -2039
  220. package/crates/team-agent/src/lifecycle/tests/core.rs +0 -1298
  221. package/crates/team-agent/src/lifecycle/tests/f032_startup_prompt_best_effort_red.rs +0 -315
  222. package/crates/team-agent/src/lifecycle/tests/fixtures/pi_list_models_g0.stdout.txt +0 -222
  223. package/crates/team-agent/src/lifecycle/tests/fixtures/pi_list_models_spaced_ids.stdout.txt +0 -9
  224. package/crates/team-agent/src/lifecycle/tests/fixtures/pi_native_mcp_host.mjs +0 -71
  225. package/crates/team-agent/src/lifecycle/tests/gate_fixtures.rs +0 -562
  226. package/crates/team-agent/src/lifecycle/tests/grok_require_explicit_model_red.rs +0 -110
  227. package/crates/team-agent/src/lifecycle/tests/harvest2_a_batch_red.rs +0 -472
  228. package/crates/team-agent/src/lifecycle/tests/host_cotenant_death_p0_contract.rs +0 -921
  229. package/crates/team-agent/src/lifecycle/tests/issue238_codex_effort_contract.rs +0 -68
  230. package/crates/team-agent/src/lifecycle/tests/lane_ops.rs +0 -1851
  231. package/crates/team-agent/src/lifecycle/tests/launch_spawn.rs +0 -3981
  232. package/crates/team-agent/src/lifecycle/tests/lifecycle_lock.rs +0 -339
  233. package/crates/team-agent/src/lifecycle/tests/lifecycle_rollback_red.rs +0 -596
  234. package/crates/team-agent/src/lifecycle/tests/main_preserved.rs +0 -513
  235. package/crates/team-agent/src/lifecycle/tests/mcp_tool_name_format_red.rs +0 -69
  236. package/crates/team-agent/src/lifecycle/tests/phase_b_contracts.rs +0 -824
  237. package/crates/team-agent/src/lifecycle/tests/phase_golden.rs +0 -723
  238. package/crates/team-agent/src/lifecycle/tests/pi_catalog_red.rs +0 -152
  239. package/crates/team-agent/src/lifecycle/tests/pi_command_red.rs +0 -182
  240. package/crates/team-agent/src/lifecycle/tests/pi_dual_entry_red.rs +0 -374
  241. package/crates/team-agent/src/lifecycle/tests/pi_executable_mcp_red.rs +0 -546
  242. package/crates/team-agent/src/lifecycle/tests/pi_lifecycle_doctor_red.rs +0 -31
  243. package/crates/team-agent/src/lifecycle/tests/pi_native_mcp_capability_red.rs +0 -528
  244. package/crates/team-agent/src/lifecycle/tests/pi_session_delivery_red.rs +0 -134
  245. package/crates/team-agent/src/lifecycle/tests/quick_start_worker_readiness_red.rs +0 -333
  246. package/crates/team-agent/src/lifecycle/tests/realmachine_clusters_1_6_red.rs +0 -786
  247. package/crates/team-agent/src/lifecycle/tests/realmachine_residual_g1_g4_red.rs +0 -648
  248. package/crates/team-agent/src/lifecycle/tests/restart.rs +0 -185
  249. package/crates/team-agent/src/lifecycle/tests/restart_build_before_destroy_0540_contract.rs +0 -1053
  250. package/crates/team-agent/src/lifecycle/tests/restart_liveness_red.rs +0 -679
  251. package/crates/team-agent/src/lifecycle/tests/restart_rebind_hotfix_252_red.rs +0 -1126
  252. package/crates/team-agent/src/lifecycle/tests/restart_session_capture_red.rs +0 -153
  253. package/crates/team-agent/src/lifecycle/tests/resume_recover_red.rs +0 -427
  254. package/crates/team-agent/src/lifecycle/tests/stale_team_saveconflict_contract.rs +0 -431
  255. package/crates/team-agent/src/lifecycle/tests/startup_latency_contract.rs +0 -1066
  256. package/crates/team-agent/src/lifecycle/tests/status_credential_redaction_contract.rs +0 -876
  257. package/crates/team-agent/src/lifecycle/tests/subprep_codex_trust_roundtrip_red.rs +0 -1026
  258. package/crates/team-agent/src/lifecycle/tests/swallow_batch4_semantics_red.rs +0 -90
  259. package/crates/team-agent/src/lifecycle/tests/team_in_team_identity_scope_red.rs +0 -543
  260. package/crates/team-agent/src/lifecycle/tests/team_in_team_sibling_quick_start_red.rs +0 -1103
  261. package/crates/team-agent/src/lifecycle/tests/team_in_team_state_scope_red.rs +0 -1276
  262. package/crates/team-agent/src/lifecycle/tests/team_key_retirement_roster_red.rs +0 -145
  263. package/crates/team-agent/src/lifecycle/tests/team_key_retirement_txn_red.rs +0 -603
  264. package/crates/team-agent/src/lifecycle/tests/test_isolation_escape_contract.rs +0 -936
  265. package/crates/team-agent/src/lifecycle/tests/upgrade_compat_0211_red.rs +0 -970
  266. package/crates/team-agent/src/lifecycle/tests/verify_rs031_window_consistency_red.rs +0 -996
  267. package/crates/team-agent/src/lifecycle/tests/worker_spawn_env_red.rs +0 -827
  268. package/crates/team-agent/src/lifecycle/tests.rs +0 -102
  269. package/crates/team-agent/src/lifecycle/types.rs +0 -861
  270. package/crates/team-agent/src/lifecycle/worker_command_context.rs +0 -370
  271. package/crates/team-agent/src/main.rs +0 -52
  272. package/crates/team-agent/src/mcp_server/helpers.rs +0 -510
  273. package/crates/team-agent/src/mcp_server/lifecycle_tools/mod.rs +0 -6
  274. package/crates/team-agent/src/mcp_server/lifecycle_tools/state_status.rs +0 -43
  275. package/crates/team-agent/src/mcp_server/mod.rs +0 -119
  276. package/crates/team-agent/src/mcp_server/normalize.rs +0 -485
  277. package/crates/team-agent/src/mcp_server/readiness.rs +0 -546
  278. package/crates/team-agent/src/mcp_server/readiness_tests.rs +0 -369
  279. package/crates/team-agent/src/mcp_server/tests/golden.rs +0 -285
  280. package/crates/team-agent/src/mcp_server/tests/normalize.rs +0 -247
  281. package/crates/team-agent/src/mcp_server/tests/scoped.rs +0 -168
  282. package/crates/team-agent/src/mcp_server/tests/send.rs +0 -388
  283. package/crates/team-agent/src/mcp_server/tests/tools.rs +0 -343
  284. package/crates/team-agent/src/mcp_server/tests/wire.rs +0 -254
  285. package/crates/team-agent/src/mcp_server/tests.rs +0 -43
  286. package/crates/team-agent/src/mcp_server/tools.rs +0 -1015
  287. package/crates/team-agent/src/mcp_server/types.rs +0 -396
  288. package/crates/team-agent/src/mcp_server/wire.rs +0 -867
  289. package/crates/team-agent/src/messaging/activity.rs +0 -461
  290. package/crates/team-agent/src/messaging/address.rs +0 -87
  291. package/crates/team-agent/src/messaging/delivery.rs +0 -3785
  292. package/crates/team-agent/src/messaging/delivery_consumption_tests.rs +0 -234
  293. package/crates/team-agent/src/messaging/enter_rescue.rs +0 -362
  294. package/crates/team-agent/src/messaging/enter_rescue_tests.rs +0 -633
  295. package/crates/team-agent/src/messaging/helpers.rs +0 -354
  296. package/crates/team-agent/src/messaging/leader_channel.rs +0 -287
  297. package/crates/team-agent/src/messaging/leader_receiver.rs +0 -786
  298. package/crates/team-agent/src/messaging/mod.rs +0 -166
  299. package/crates/team-agent/src/messaging/persist.rs +0 -377
  300. package/crates/team-agent/src/messaging/presentation.rs +0 -417
  301. package/crates/team-agent/src/messaging/results/collection_tests.rs +0 -719
  302. package/crates/team-agent/src/messaging/results.rs +0 -1415
  303. package/crates/team-agent/src/messaging/scheduler.rs +0 -217
  304. package/crates/team-agent/src/messaging/selftest.rs +0 -315
  305. package/crates/team-agent/src/messaging/send.rs +0 -981
  306. package/crates/team-agent/src/messaging/tests/basic.rs +0 -346
  307. package/crates/team-agent/src/messaging/tests/dup_inject.rs +0 -551
  308. package/crates/team-agent/src/messaging/tests/e23.rs +0 -49
  309. package/crates/team-agent/src/messaging/tests/leader_channel.rs +0 -169
  310. package/crates/team-agent/src/messaging/tests/leader_inject_acceptance.rs +0 -817
  311. package/crates/team-agent/src/messaging/tests/main_preserved.rs +0 -85
  312. package/crates/team-agent/src/messaging/tests/mod.rs +0 -306
  313. package/crates/team-agent/src/messaging/tests/runtime.rs +0 -3655
  314. package/crates/team-agent/src/messaging/tests/spine.rs +0 -837
  315. package/crates/team-agent/src/messaging/tests/wave2.rs +0 -381
  316. package/crates/team-agent/src/messaging/trust.rs +0 -215
  317. package/crates/team-agent/src/messaging/types.rs +0 -471
  318. package/crates/team-agent/src/messaging/wait.rs +0 -366
  319. package/crates/team-agent/src/messaging/watchers.rs +0 -916
  320. package/crates/team-agent/src/model/enums.rs +0 -356
  321. package/crates/team-agent/src/model/errors.rs +0 -18
  322. package/crates/team-agent/src/model/ids.rs +0 -165
  323. package/crates/team-agent/src/model/mod.rs +0 -29
  324. package/crates/team-agent/src/model/name_similarity.rs +0 -252
  325. package/crates/team-agent/src/model/pane_authority_refusal.rs +0 -359
  326. package/crates/team-agent/src/model/paths.rs +0 -257
  327. package/crates/team-agent/src/model/routing.rs +0 -387
  328. package/crates/team-agent/src/model/spec.rs +0 -1184
  329. package/crates/team-agent/src/model/task_graph.rs +0 -321
  330. package/crates/team-agent/src/model/testdata/fuzz.golden.yaml +0 -43
  331. package/crates/team-agent/src/model/testdata/fuzz.yaml +0 -43
  332. package/crates/team-agent/src/model/testdata/spec_invalid_a.yaml +0 -209
  333. package/crates/team-agent/src/model/testdata/team.spec.golden.yaml +0 -208
  334. package/crates/team-agent/src/model/testdata/team.spec.yaml +0 -208
  335. package/crates/team-agent/src/model/yaml/tests.rs +0 -188
  336. package/crates/team-agent/src/model/yaml.rs +0 -802
  337. package/crates/team-agent/src/os_probe.rs +0 -280
  338. package/crates/team-agent/src/packaging/install.rs +0 -219
  339. package/crates/team-agent/src/packaging/migrate.rs +0 -38
  340. package/crates/team-agent/src/packaging/mod.rs +0 -91
  341. package/crates/team-agent/src/packaging/repair.rs +0 -33
  342. package/crates/team-agent/src/packaging/tests.rs +0 -822
  343. package/crates/team-agent/src/packaging/types.rs +0 -417
  344. package/crates/team-agent/src/platform/argv.rs +0 -331
  345. package/crates/team-agent/src/platform/errors.rs +0 -104
  346. package/crates/team-agent/src/platform/file_lock.rs +0 -279
  347. package/crates/team-agent/src/platform/mod.rs +0 -5
  348. package/crates/team-agent/src/platform/process.rs +0 -545
  349. package/crates/team-agent/src/provider/adapter.rs +0 -898
  350. package/crates/team-agent/src/provider/adapters/copilot.rs +0 -75
  351. package/crates/team-agent/src/provider/adapters/copilot_fork.rs +0 -307
  352. package/crates/team-agent/src/provider/adapters/fake.rs +0 -18
  353. package/crates/team-agent/src/provider/adapters/mod.rs +0 -28
  354. package/crates/team-agent/src/provider/approvals/mod.rs +0 -3
  355. package/crates/team-agent/src/provider/approvals/parsing.rs +0 -586
  356. package/crates/team-agent/src/provider/approvals/runtime_prompts.rs +0 -205
  357. package/crates/team-agent/src/provider/argv_route.rs +0 -405
  358. package/crates/team-agent/src/provider/bypass_flags.rs +0 -116
  359. package/crates/team-agent/src/provider/classify.rs +0 -773
  360. package/crates/team-agent/src/provider/command_helpers.rs +0 -45
  361. package/crates/team-agent/src/provider/faults.rs +0 -165
  362. package/crates/team-agent/src/provider/helpers.rs +0 -52
  363. package/crates/team-agent/src/provider/mod.rs +0 -78
  364. package/crates/team-agent/src/provider/model_catalog.rs +0 -722
  365. package/crates/team-agent/src/provider/session/capture.rs +0 -2631
  366. package/crates/team-agent/src/provider/session/context_fork/outcome.rs +0 -168
  367. package/crates/team-agent/src/provider/session/context_fork.rs +0 -264
  368. package/crates/team-agent/src/provider/session/mod.rs +0 -36
  369. package/crates/team-agent/src/provider/session/resume.rs +0 -269
  370. package/crates/team-agent/src/provider/session_scan/common/tests.rs +0 -316
  371. package/crates/team-agent/src/provider/session_scan/common.rs +0 -618
  372. package/crates/team-agent/src/provider/session_scan/copilot.rs +0 -280
  373. package/crates/team-agent/src/provider/session_scan/pi.rs +0 -152
  374. package/crates/team-agent/src/provider/session_scan.rs +0 -61
  375. package/crates/team-agent/src/provider/startup_prompt.rs +0 -757
  376. package/crates/team-agent/src/provider/submit_now.rs +0 -71
  377. package/crates/team-agent/src/provider/testdata/copilot-ready-marker.txt +0 -5
  378. package/crates/team-agent/src/provider/testdata/copilot-trust-prompt.txt +0 -19
  379. package/crates/team-agent/src/provider/testdata/grok-models-authenticated.stdout +0 -9
  380. package/crates/team-agent/src/provider/testdata/grok-models-unauthenticated.stdout +0 -7
  381. package/crates/team-agent/src/provider/tests/adapter.rs +0 -354
  382. package/crates/team-agent/src/provider/tests/classify.rs +0 -201
  383. package/crates/team-agent/src/provider/tests/command_helpers.rs +0 -26
  384. package/crates/team-agent/src/provider/tests/copilot_fork.rs +0 -191
  385. package/crates/team-agent/src/provider/tests/faults.rs +0 -200
  386. package/crates/team-agent/src/provider/tests/idle.rs +0 -140
  387. package/crates/team-agent/src/provider/tests/wire.rs +0 -213
  388. package/crates/team-agent/src/provider/tests.rs +0 -32
  389. package/crates/team-agent/src/provider/types.rs +0 -529
  390. package/crates/team-agent/src/provider/wire.rs +0 -406
  391. package/crates/team-agent/src/redaction.rs +0 -259
  392. package/crates/team-agent/src/state/abnormal_watch/tests/mod.rs +0 -248
  393. package/crates/team-agent/src/state/abnormal_watch.rs +0 -144
  394. package/crates/team-agent/src/state/identity.rs +0 -914
  395. package/crates/team-agent/src/state/identity_keys.rs +0 -132
  396. package/crates/team-agent/src/state/mod.rs +0 -74
  397. package/crates/team-agent/src/state/owner_gate.rs +0 -482
  398. package/crates/team-agent/src/state/ownership.rs +0 -563
  399. package/crates/team-agent/src/state/paths.rs +0 -395
  400. package/crates/team-agent/src/state/persist.rs +0 -3811
  401. package/crates/team-agent/src/state/projection.rs +0 -1508
  402. package/crates/team-agent/src/state/repository/intent.rs +0 -14
  403. package/crates/team-agent/src/state/repository/tests.rs +0 -184
  404. package/crates/team-agent/src/state/repository.rs +0 -866
  405. package/crates/team-agent/src/state/selector.rs +0 -160
  406. package/crates/team-agent/src/state/testdata/state-rich.canonical.json +0 -133
  407. package/crates/team-agent/src/tmux_backend/enter_rescue.rs +0 -325
  408. package/crates/team-agent/src/tmux_backend/enter_rescue_tests.rs +0 -432
  409. package/crates/team-agent/src/tmux_backend/native_input.rs +0 -513
  410. package/crates/team-agent/src/tmux_backend/native_input_tests.rs +0 -742
  411. package/crates/team-agent/src/tmux_backend/tests.rs +0 -4254
  412. package/crates/team-agent/src/tmux_backend.rs +0 -4527
  413. package/crates/team-agent/src/topology.rs +0 -537
  414. package/crates/team-agent/src/transport/enter_rescue.rs +0 -166
  415. package/crates/team-agent/src/transport/native_input.rs +0 -491
  416. package/crates/team-agent/src/transport/test_support.rs +0 -743
  417. package/crates/team-agent/src/transport/tests/behavior.rs +0 -326
  418. package/crates/team-agent/src/transport/tests/mod.rs +0 -194
  419. package/crates/team-agent/src/transport/tests/wire.rs +0 -574
  420. package/crates/team-agent/src/transport.rs +0 -1655
  421. package/crates/team-agent/src/transport_factory.rs +0 -819
@@ -10,6 +10,12 @@ The current daily forms below are coverage authorities, not permission to execut
10
10
  ### Current documented forms
11
11
  ```text
12
12
  team-agent --version
13
+ team-agent claim-leader --confirm
14
+ team-agent pi
15
+ team-agent start
16
+ team-agent start AGENT --provider PROVIDER --model MODEL --session fresh
17
+ team-agent start reviewer
18
+ team-agent stop <agent>
13
19
  team-agent agent clone reviewer --as second-reviewer --workspace . --team project-team
14
20
  team-agent agent fork reviewer --as continued-reviewer --workspace . --team project-team
15
21
  team-agent agent remove reviewer --confirm --workspace . --team project-team
@@ -33,11 +39,14 @@ team-agent route show pi --json
33
39
  team-agent route status --json
34
40
  team-agent send reviewer "Review the current change and report the main risk." --workspace . --team project-team
35
41
  team-agent send reviewer "Review this change" --workspace . --team project-team
42
+ team-agent send worker '/compact' --bare --workspace . --team project-team
43
+ team-agent send --targets worker '/compact' --bare --workspace . --team project-team
44
+ team-agent send worker --bare --workspace . --team project-team -- '/compact retain unresolved issues'
36
45
  team-agent start --from ./my-team --workspace . --team project-team
37
46
  team-agent start --from .team/current --workspace . --team project-team
38
47
  team-agent start --help
39
48
  team-agent start --workspace . --team project-team
40
- team-agent start analyst --provider pi --model openai-codex/gpt-6-luna --bypass false --prompt "Analyze the proposed change"
49
+ team-agent start analyst --provider pi --bypass false --prompt "Analyze the proposed change"
41
50
  team-agent start auditor --role-file /absolute/path/to/auditor.md
42
51
  team-agent start reviewer --effort high --bypass false
43
52
  team-agent start reviewer --effort high --bypass false --workspace . --team project-team
@@ -52,23 +61,22 @@ team-agent team restart --workspace . --team project-team
52
61
  team-agent team restart --workspace . --team project-team --json
53
62
  ```
54
63
 
55
- ### Retained compatibility and negative-probe forms — NOT daily recipes
56
- The following old argv remain in the closed-world catalog for existing scripts and historical tests. Some intentionally invalid or removed flag combinations are negative probes, not supported operational advice. Existing covered records keep their original evidence and are not renamed to imply new-command coverage.
64
+ ### Supplemental canonical forms and negative probes — NOT blanket execution authorization
65
+ The following forms supplement the daily recipes. Intentionally invalid flag combinations remain negative probes, not supported operational advice. Historical compatibility evidence is not renamed to claim coverage for canonical spellings; new forms require their own evidence.
57
66
 
58
67
  ```text
59
- team-agent add-agent <agent> --role-file <file>
60
- team-agent add-agent reviewer --role-file .team/current/agents/reviewer.md --workspace .
68
+ team-agent start <agent> --role-file <file>
69
+ team-agent start reviewer --role-file .team/current/agents/reviewer.md --workspace .
61
70
  team-agent approvals
62
71
  team-agent approvals <agent_id>
63
72
  team-agent approvals [coder]
64
- team-agent attach-leader
65
73
  team-agent claim-leader
66
74
  team-agent claude
67
- team-agent clone-agent <source> --as <new>
75
+ team-agent agent clone <source> --as <new>
68
76
  team-agent codex
69
77
  team-agent codex --dangerously-bypass-approvals-and-sandbox
70
78
  team-agent doctor
71
- team-agent fork-agent <source> --as <new>
79
+ team-agent agent fork <source> --as <new>
72
80
  team-agent inbox <agent_id> -n 3
73
81
  team-agent inbox coder -n 3
74
82
  team-agent profile doctor <name> --workspace . --json
@@ -78,14 +86,14 @@ team-agent profile init codex-default --auth-mode subscription --workspace .
78
86
  team-agent profile init deepseek --auth-mode compatible_api --workspace .
79
87
  team-agent profile show <name> --workspace . --json
80
88
  team-agent profile show deepseek --workspace . --json
81
- team-agent quick-start
82
- team-agent quick-start ./roles --team-id alpha
83
- team-agent quick-start .team/alpha
84
- team-agent quick-start .team/current
85
- team-agent quick-start <dir>
86
- team-agent quick-start <plain-text-team-dir>
87
- team-agent remove-agent <agent> --workspace . --confirm
88
- team-agent reset-agent <agent> --discard-session
89
+ team-agent start --from .
90
+ team-agent start --from ./roles --team alpha
91
+ team-agent start --from .team/alpha
92
+ team-agent start --from .team/current
93
+ team-agent start --from <dir>
94
+ team-agent start --from <plain-text-team-dir>
95
+ team-agent agent remove <agent> --workspace . --confirm
96
+ team-agent agent reset <agent> --discard-session
89
97
  team-agent restart
90
98
  team-agent restart .
91
99
  team-agent restart . --allow-fresh
@@ -96,10 +104,10 @@ team-agent send --watch-result coder "Do the bounded task"
96
104
  team-agent send TO MESSAGE
97
105
  team-agent send reviewer "..."
98
106
  team-agent send reviewer "Review this change"
99
- team-agent shutdown --workspace . --keep-logs
100
- team-agent start-agent <agent>
101
- team-agent start-agent <agent_id> --workspace .
102
- team-agent start-agent coder --workspace .
107
+ team-agent stop --workspace .
108
+ team-agent start <agent>
109
+ team-agent start <agent_id> --workspace .
110
+ team-agent start coder --workspace .
103
111
  team-agent status
104
112
  team-agent status --detail --json
105
113
  team-agent status coder
@@ -130,7 +138,7 @@ team-agent inbox leader --workspace . --team project-team -n 3
130
138
  team-agent stop --workspace . --team project-team --json
131
139
  ```
132
140
 
133
- Creation needs complete configuration; start is not automatic template generation or state repair. Later start starts stopped non-paused members, including explicitly stopped members, and leaves verified live/drift-free instances unchanged. A no-op does not synchronize role bodies, apply new launch settings or prove tool/business readiness. A queued send is not completion; wait for a natural reply/result.
141
+ `start AGENT` can create an ad-hoc seat without pre-existing files, or adopt its existing role document. Team creation from `--from` uses a complete source. Later team start preserves explicitly stopped/paused members and leaves verified live seats unchanged; `start AGENT` explicitly reactivates one seat. Start is not permission to repair corrupt ownership. A no-op does not synchronize role bodies, apply new launch settings or prove tool/business readiness. A queued send is not completion; wait for a natural reply/result.
134
142
 
135
143
  ## Scope and deterministic targets
136
144
 
@@ -142,9 +150,9 @@ Creation needs complete configuration; start is not automatic template generatio
142
150
 
143
151
  ## User decisions and compatibility
144
152
 
145
- Do not silently pick provider, model, effort or bypass. False preserves native permissions; it is not a read-only or sandbox guarantee. Use supported `models --provider NAME` discovery and retain exact model IDs.
153
+ Honor explicit provider, model, effort and bypass choices. The unified resolver reports omitted decisions in `defaulted`; provider defaults to a supported leader provider or Pi, bypass to false, and native model/effort defaults are used only where the provider admits them. False preserves native permissions; it is not a read-only or sandbox guarantee. Use supported `models --provider NAME` discovery and retain exact model IDs.
146
154
 
147
- Traditional minimal Markdown remains accepted by the original compiler. Its compatibility interpretation is **not a default for a new user plan**: provider defaults to Pi, bypass to false, and unset effort is left to the provider. Claude, Codex, Grok and Pi workers run only through their contract leaves, which require an exact model ID from `team-agent models --provider claude|codex|grok|pi`; an unset model is refused, never guessed. Team Agent does not infer a role model from TEAM defaults. Contract leaves use only an explicit role effort; legacy explicit TEAM provider_effort remains supported for the remaining legacy roles (Copilot, Gemini); do not invent an effort for unspecified roles.
155
+ Traditional minimal Markdown remains accepted by the original compiler. Omitted values use the same safe declaration defaults: provider falls back to Pi, bypass to false, and unset model/effort are left to the provider where admitted. Codex and Grok workers run only through their contract leaves, which require an exact model ID from `team-agent models --provider codex|grok`; an unset model is refused, never guessed. Team Agent does not infer a role model from TEAM defaults. Contract leaves use only an explicit role effort; legacy explicit TEAM provider_effort remains supported for non-Pi legacy roles (Claude, Copilot, Gemini); do not invent an effort for unspecified roles.
148
156
 
149
157
  Supported role metadata:
150
158
  - `agent_id` (legacy `name`) and `role` identify the seat.
@@ -155,16 +163,16 @@ Supported role metadata:
155
163
 
156
164
  ## Add or wake with start
157
165
 
158
- A RoleOnly seat can be registered from its complete role file. A genuinely absent ID needs a complete file or explicit provider/bypass choices. These Pi/false values are examples of user decisions, not defaults:
166
+ One `start AGENT` contract creates absent seats, adopts existing role files, and resumes stopped seats. No prewritten file or repeated provider/bypass decision is required for an ad-hoc start. Receipts include `session_policy`, an effective `declaration`, and the `defaulted` field list:
159
167
 
160
168
  ```sh
161
169
  team-agent start auditor --role-file /absolute/path/to/auditor.md
162
- team-agent start analyst --provider pi --model openai-codex/gpt-6-luna --bypass false --prompt "Analyze the proposed change"
170
+ team-agent start analyst --provider pi --bypass false --prompt "Analyze the proposed change"
163
171
  ```
164
172
 
165
- An imported external role goes into this team's `agents/ID.md`; the original stays untouched. A file already at that destination is used in place. Equal file/CLI definitions are accepted; conflicting provider/model/effort/bypass/prompt/profile values refuse before writes/startup. Omitted model, effort, profile and role body remain unchanged.
173
+ An imported external role goes into this team's `agents/ID.md`; the original stays untouched. A file already at that destination is used in place. Explicit CLI fields override the selected declaration. Omitted existing fields and role instructions remain unchanged; genuinely omitted defaults are identified in receipts.
166
174
 
167
- Use `agent add` when explicitly requesting strict add-only behavior. It creates and starts a new seat but refuses an occupied ID, including stopped seats. Its disaster-recovery `--force` still requires a positively dead registered old pane and no live target cohort; a live or unknown target refuses. Force is not a role-update or routine restart shortcut.
175
+ Hidden spelling aliases delegate directly to `start AGENT`. There is no second strict-add contract and no start `--force` option. Unknown ownership or liveness refuses rather than granting permission to replace a process.
168
176
 
169
177
  ## Change a registered seat
170
178
 
@@ -173,9 +181,9 @@ team-agent stop reviewer --workspace . --team project-team
173
181
  team-agent start reviewer --effort high --bypass false --workspace . --team project-team
174
182
  ```
175
183
 
176
- Stop first before applying edited instructions or launch settings. A running seat with requested changes or known drift refuses rather than claiming a successful no-op. `--prompt` replaces permanent role instructions, not a one-time task; send one-time work with `send`. `--bypass` requires explicit true/false. Matching provider is allowed, but changing engines, even by file edits, is refused; use a new ID.
184
+ Stop first before applying edited instructions or launch settings. A running seat with requested changes or known drift refuses rather than claiming a successful no-op. `--prompt` replaces permanent role instructions, not a one-time task; send one-time work with `send`. `--bypass` requires explicit true/false. A stopped seat may change provider/model with `start AGENT --provider PROVIDER --model MODEL --session fresh`. Its logical ID and history are retained. Provider changes under preserve or auto refuse and instruct fresh. A live seat with identical configuration is a no-op; any changed configuration or fresh-session request requires `stop AGENT` first.
177
185
 
178
- The original role/spec transaction and provider admission remain authoritative. A failure is not an applied update; never assume a partial/unknown native replacement was rolled back or is safe to retry. Inspect the exact scoped receipt, status and doctor first.
186
+ The original role/spec transaction and provider admission remain authoritative. A failure can follow a committed declaration or uncertain launch; never assume a partial/unknown native replacement was rolled back or is safe to retry. Inspect the exact scoped receipt, status and doctor first.
179
187
 
180
188
  ## Restart and session policy
181
189
 
@@ -187,9 +195,11 @@ team-agent agent restart reviewer --session preserve --workspace . --team projec
187
195
  team-agent team restart --session preserve --workspace . --team project-team
188
196
  ```
189
197
 
190
- New agent/team restart defaults to **auto**: proven resume when possible, otherwise only the provider's existing legal fresh path. Auto is not a guarantee of saved memory, reset permission or a fallback around native capability refusal. Preserve requests strict continuation. The old bare/explicit-path restart bridge retains its original preserve default and output.
198
+ New agent/team restart defaults to **auto**: proven resume when possible, otherwise only the provider's existing legal fresh path. Auto is not a guarantee of saved memory, reset permission or a fallback around native capability refusal. Preserve requests strict continuation. Start defaults to preserve for existing resumable context. Contextless seats and legacy providers without resume support start fresh and report `op=fresh`; launch timestamps and pending IDs alone do not require continuation. `start AGENT --session fresh` explicitly creates a new native session while retaining old history. Fresh and `--allow-fresh` cannot be combined. The old bare/explicit-path restart bridge retains its original preserve default and output.
191
199
 
192
- Single-agent restart checks scope, role/session backing and capabilities before stop; only confirmed stop allows start. On partial failure, report actual completed steps and inspect; do not blindly reset or replay. Team restart does not reactivate explicitly stopped or paused seats. Whole-team start intentionally activates the selected non-paused roster; neither command silently unpauses.
200
+ Single-agent restart checks scope, role/session backing and capabilities before stop; only confirmed stop allows start. On partial failure, report actual completed steps and inspect; do not blindly reset or replay. Team restart does not reactivate explicitly stopped or paused seats. Whole-team start also preserves explicit stops; use `start AGENT` for an intentional wake-up. Neither command silently unpauses.
201
+
202
+ Ordinary framed managed sends after a cold launch wait up to 60 seconds for initialization evidence belonging to that launch before injecting input. A readiness timeout leaves the message stored but injects nothing; inspect status/doctor before retrying. Mailbox-only delivery does not wait. Completed startup evidence survives helper exit or binary upgrades: this is not an ongoing MCP-health gate, nor proof of business acceptance.
193
203
 
194
204
  ## Explicit agent operations
195
205
 
@@ -213,12 +223,36 @@ team-agent profile init my-api --auth-mode compatible_api --workspace .
213
223
 
214
224
  A role may explicitly reference `auth_mode: compatible_api` and `profile: my-api`, with the user's chosen provider/bypass. Profile model applies only when the role has no explicit override. Subscription and custom-API capabilities differ; follow actual provider admission, not a generic promise.
215
225
 
226
+ ## Native REPL input: `--bare`
227
+
228
+ `--bare` removes message framing, not addressing safeguards. It is only for a trusted leader/operator sending one live, roster-bound non-leader worker a native input line. It is not exposed by MCP. Ordinary tasks must continue to use framed send.
229
+
230
+ ```sh
231
+ team-agent send worker '/compact' --bare --workspace . --team project-team
232
+ team-agent send --targets worker '/compact' --bare --workspace . --team project-team
233
+ team-agent send worker --bare --workspace . --team project-team -- '/compact retain unresolved issues'
234
+ ```
235
+
236
+ Only Claude, Codex and Pi are supported, including their existing canonical aliases. The content must be nonblank, one line, at most 512 UTF-8 bytes, without control characters or Unicode line separators. Admitted bytes, including surrounding spaces and a trailing semicolon, are preserved. The `--` terminator makes every following token payload, even `--socket`, `--json` or `--bare`. No `--raw` alias or message-file source is added.
237
+
238
+ Bare refuses mailbox delivery, leader targets, fanout, unmanaged `--socket`/`--pane` addressing, `--dialect` overrides, and `--task`, `--presentation-sink`, `--message-class`, `--case-id`. Use the existing comma-based `--targets worker` syntax, not a JSON array. The lifecycle commands `/exit` and `/quit` are refused; use `team-agent stop <agent>` instead.
239
+
240
+ Physical delivery requires an empty, idle native composer and verified target ownership. A busy/draft/modal/unknown surface is refused without clearing user input. Claude and Codex use unbracketed `paste-buffer -r`; Pi uses `-p -r` to avoid slash autocomplete consuming Enter. The original payload has no heading, token or Codex newline trailer; after a 100 ms floor, the owned live composer is checked and exactly one native Enter is sent. Never treat shell metacharacters as shell code.
241
+
242
+ The existing inbox JSONL remains the audit source: original content, sender, target and submission evidence are retained with `framing: "bare"`. No `events.jsonl` is created. Explicit message ids retain duplicate protection; changing framing under the same id yields `request_id_conflict`. Anonymous sends remain separate invocations. After a post-paste uncertainty, do not resend or clear the composer automatically.
243
+
244
+ `accepted: null` remains authoritative: physical submission is not proof of native-command completion. Bare does not update `first_send_at`, task attribution or session identity. A success receipt reports `session_binding: "not_reobserved"`; a session-changing command such as `/clear` may leave the previous observed session id stale. Use lifecycle-managed reset/start for state-tracked session changes; bare does not guarantee later resume continuity.
245
+
216
246
  ## Observe and finish
217
247
 
248
+ **Do not poll.** Never run `inbox`, `status` or `doctor` repeatedly, in sleep/retry loops, or as repeated tool calls. Live messages are pushed directly to terminals. Finish the current work, stop issuing tool calls, and wait for a new incoming message. Run a deliberate inspection again only after a new event, relevant state-changing action, or explicit user request; an empty inbox or busy status is not a reason. Mailbox-only messages require deliberate history review.
249
+
250
+ These three commands append one `[Advisory: do not poll]` block to human output, on the existing output stream (including stderr for human errors). With `--json`, the same guidance is the root string `polling_advisory`: one valid JSON document, never trailing non-JSON text. Existing fields, nested boolean `advisory` values and exit codes retain their meanings; strict schemas may need to allow the additional root key.
251
+
218
252
  `status` / `team status` expose the concise node projection; `--detail` does not restore historical internal dumps. Use `doctor` for diagnosis. Check actual readiness and returned recovery actions; unknown evidence is not a stopped/live proof. Stop with `stop --workspace PATH --team KEY --json` and verify scoped residual sessions, processes and owned files. Protect parent/sibling/shared resources.
219
253
 
220
254
  Workers have only `send_message`, `report_result` and optional `get_team_status`. Report completion through `report_result`; do not promote a worker to lifecycle operator. See [nested teams](https://github.com/Florious95/team-agent/blob/main/skills/team-agent/references/team-in-team.md) and [recovery](https://github.com/Florious95/team-agent/blob/main/skills/team-agent/references/recovery-runbook.md).
221
255
 
222
256
  ## Legacy command bridges
223
257
 
224
- Existing scripts may retain `quick-start`, `shutdown`, and `add-agent/start-agent/stop-agent/remove-agent/reset-agent/clone-agent/fork-agent`. These are compatibility interfaces, not the daily workflow. Use the short/team/agent forms above for new instructions. Do not mechanically rename scripts: legacy start refuses a live seat, new start can prove a no-op; bare/explicit-path legacy restart preserves its default. Old normal command output and failure evidence remain authoritative.
258
+ Existing scripts may retain hidden compatibility spellings, but new instructions use the canonical start/stop/team/agent forms above. The start aliases share one resolver, defaults, session policy, no-op/refusal rules and receipts. Bare/explicit-path legacy restart preserves its separate documented default. Startup evidence remains authoritative.
@@ -106,6 +106,6 @@ Warn that context will be lost. Each intentional fresh/reset must retain `recove
106
106
 
107
107
  ## Compatibility and validation scope
108
108
 
109
- The old bare/explicit-path `restart` keeps its preserve default; `quick-start`, `shutdown` and suffixed agent commands remain compatibility bridges for existing scripts, not these daily recovery instructions. Do not mechanically rename scripts or rewrite historical receipts as if they used the new surface.
109
+ The old bare/explicit-path `restart` keeps its preserve default. Hidden compatibility spellings remain available for existing scripts; daily recovery uses canonical `team-agent start`, `team-agent stop` and `team-agent agent` operations. Do not mechanically rename scripts or rewrite historical receipts as if they used the new surface.
110
110
 
111
111
  The short start/stop/restart journey has isolated two-Pi lifecycle verification, including live no-op, held-seat preservation, both-seat activation and scoped cleanup. Automated fixtures cover continuation/backing/partial/native/mixed refusals. This is not a fresh all-provider same-session memory or all-platform acceptance claim. Earlier Rust resume/cleanup records retain their original scope in history; Python-only repair, roster cloning, raw state injection and direct server/coordinator surgery are not promoted into validated public recipes.
@@ -10,6 +10,12 @@ The current daily forms below are coverage authorities, not permission to execut
10
10
  ### Current documented forms
11
11
  ```text
12
12
  team-agent --version
13
+ team-agent claim-leader --confirm
14
+ team-agent pi
15
+ team-agent start
16
+ team-agent start AGENT --provider PROVIDER --model MODEL --session fresh
17
+ team-agent start reviewer
18
+ team-agent stop <agent>
13
19
  team-agent agent clone reviewer --as second-reviewer --workspace . --team project-team
14
20
  team-agent agent fork reviewer --as continued-reviewer --workspace . --team project-team
15
21
  team-agent agent remove reviewer --confirm --workspace . --team project-team
@@ -33,11 +39,14 @@ team-agent route show pi --json
33
39
  team-agent route status --json
34
40
  team-agent send reviewer "Review the current change and report the main risk." --workspace . --team project-team
35
41
  team-agent send reviewer "Review this change" --workspace . --team project-team
42
+ team-agent send worker '/compact' --bare --workspace . --team project-team
43
+ team-agent send --targets worker '/compact' --bare --workspace . --team project-team
44
+ team-agent send worker --bare --workspace . --team project-team -- '/compact retain unresolved issues'
36
45
  team-agent start --from ./my-team --workspace . --team project-team
37
46
  team-agent start --from .team/current --workspace . --team project-team
38
47
  team-agent start --help
39
48
  team-agent start --workspace . --team project-team
40
- team-agent start analyst --provider pi --model openai-codex/gpt-6-luna --bypass false --prompt "Analyze the proposed change"
49
+ team-agent start analyst --provider pi --bypass false --prompt "Analyze the proposed change"
41
50
  team-agent start auditor --role-file /absolute/path/to/auditor.md
42
51
  team-agent start reviewer --effort high --bypass false
43
52
  team-agent start reviewer --effort high --bypass false --workspace . --team project-team
@@ -52,23 +61,22 @@ team-agent team restart --workspace . --team project-team
52
61
  team-agent team restart --workspace . --team project-team --json
53
62
  ```
54
63
 
55
- ### Retained compatibility and negative-probe forms — NOT daily recipes
56
- The following old argv remain in the closed-world catalog for existing scripts and historical tests. Some intentionally invalid or removed flag combinations are negative probes, not supported operational advice. Existing covered records keep their original evidence and are not renamed to imply new-command coverage.
64
+ ### Supplemental canonical forms and negative probes — NOT blanket execution authorization
65
+ The following forms supplement the daily recipes. Intentionally invalid flag combinations remain negative probes, not supported operational advice. Historical compatibility evidence is not renamed to claim coverage for canonical spellings; new forms require their own evidence.
57
66
 
58
67
  ```text
59
- team-agent add-agent <agent> --role-file <file>
60
- team-agent add-agent reviewer --role-file .team/current/agents/reviewer.md --workspace .
68
+ team-agent start <agent> --role-file <file>
69
+ team-agent start reviewer --role-file .team/current/agents/reviewer.md --workspace .
61
70
  team-agent approvals
62
71
  team-agent approvals <agent_id>
63
72
  team-agent approvals [coder]
64
- team-agent attach-leader
65
73
  team-agent claim-leader
66
74
  team-agent claude
67
- team-agent clone-agent <source> --as <new>
75
+ team-agent agent clone <source> --as <new>
68
76
  team-agent codex
69
77
  team-agent codex --dangerously-bypass-approvals-and-sandbox
70
78
  team-agent doctor
71
- team-agent fork-agent <source> --as <new>
79
+ team-agent agent fork <source> --as <new>
72
80
  team-agent inbox <agent_id> -n 3
73
81
  team-agent inbox coder -n 3
74
82
  team-agent profile doctor <name> --workspace . --json
@@ -78,14 +86,14 @@ team-agent profile init codex-default --auth-mode subscription --workspace .
78
86
  team-agent profile init deepseek --auth-mode compatible_api --workspace .
79
87
  team-agent profile show <name> --workspace . --json
80
88
  team-agent profile show deepseek --workspace . --json
81
- team-agent quick-start
82
- team-agent quick-start ./roles --team-id alpha
83
- team-agent quick-start .team/alpha
84
- team-agent quick-start .team/current
85
- team-agent quick-start <dir>
86
- team-agent quick-start <plain-text-team-dir>
87
- team-agent remove-agent <agent> --workspace . --confirm
88
- team-agent reset-agent <agent> --discard-session
89
+ team-agent start --from .
90
+ team-agent start --from ./roles --team alpha
91
+ team-agent start --from .team/alpha
92
+ team-agent start --from .team/current
93
+ team-agent start --from <dir>
94
+ team-agent start --from <plain-text-team-dir>
95
+ team-agent agent remove <agent> --workspace . --confirm
96
+ team-agent agent reset <agent> --discard-session
89
97
  team-agent restart
90
98
  team-agent restart .
91
99
  team-agent restart . --allow-fresh
@@ -96,10 +104,10 @@ team-agent send --watch-result coder "Do the bounded task"
96
104
  team-agent send TO MESSAGE
97
105
  team-agent send reviewer "..."
98
106
  team-agent send reviewer "Review this change"
99
- team-agent shutdown --workspace . --keep-logs
100
- team-agent start-agent <agent>
101
- team-agent start-agent <agent_id> --workspace .
102
- team-agent start-agent coder --workspace .
107
+ team-agent stop --workspace .
108
+ team-agent start <agent>
109
+ team-agent start <agent_id> --workspace .
110
+ team-agent start coder --workspace .
103
111
  team-agent status
104
112
  team-agent status --detail --json
105
113
  team-agent status coder
@@ -130,7 +138,7 @@ team-agent inbox leader --workspace . --team project-team -n 3
130
138
  team-agent stop --workspace . --team project-team --json
131
139
  ```
132
140
 
133
- Creation needs complete configuration; start is not automatic template generation or state repair. Later start starts stopped non-paused members, including explicitly stopped members, and leaves verified live/drift-free instances unchanged. A no-op does not synchronize role bodies, apply new launch settings or prove tool/business readiness. A queued send is not completion; wait for a natural reply/result.
141
+ `start AGENT` can create an ad-hoc seat without pre-existing files, or adopt its existing role document. Team creation from `--from` uses a complete source. Later team start preserves explicitly stopped/paused members and leaves verified live seats unchanged; `start AGENT` explicitly reactivates one seat. Start is not permission to repair corrupt ownership. A no-op does not synchronize role bodies, apply new launch settings or prove tool/business readiness. A queued send is not completion; wait for a natural reply/result.
134
142
 
135
143
  ## Scope and deterministic targets
136
144
 
@@ -142,9 +150,9 @@ Creation needs complete configuration; start is not automatic template generatio
142
150
 
143
151
  ## User decisions and compatibility
144
152
 
145
- Do not silently pick provider, model, effort or bypass. False preserves native permissions; it is not a read-only or sandbox guarantee. Use supported `models --provider NAME` discovery and retain exact model IDs.
153
+ Honor explicit provider, model, effort and bypass choices. The unified resolver reports omitted decisions in `defaulted`; provider defaults to a supported leader provider or Pi, bypass to false, and native model/effort defaults are used only where the provider admits them. False preserves native permissions; it is not a read-only or sandbox guarantee. Use supported `models --provider NAME` discovery and retain exact model IDs.
146
154
 
147
- Traditional minimal Markdown remains accepted by the original compiler. Its compatibility interpretation is **not a default for a new user plan**: provider defaults to Pi, bypass to false, and unset effort is left to the provider. Claude, Codex, Grok and Pi workers run only through their contract leaves, which require an exact model ID from `team-agent models --provider claude|codex|grok|pi`; an unset model is refused, never guessed. Team Agent does not infer a role model from TEAM defaults. Contract leaves use only an explicit role effort; legacy explicit TEAM provider_effort remains supported for the remaining legacy roles (Copilot, Gemini); do not invent an effort for unspecified roles.
155
+ Traditional minimal Markdown remains accepted by the original compiler. Omitted values use the same safe declaration defaults: provider falls back to Pi, bypass to false, and unset model/effort are left to the provider where admitted. Codex and Grok workers run only through their contract leaves, which require an exact model ID from `team-agent models --provider codex|grok`; an unset model is refused, never guessed. Team Agent does not infer a role model from TEAM defaults. Contract leaves use only an explicit role effort; legacy explicit TEAM provider_effort remains supported for non-Pi legacy roles (Claude, Copilot, Gemini); do not invent an effort for unspecified roles.
148
156
 
149
157
  Supported role metadata:
150
158
  - `agent_id` (legacy `name`) and `role` identify the seat.
@@ -155,16 +163,16 @@ Supported role metadata:
155
163
 
156
164
  ## Add or wake with start
157
165
 
158
- A RoleOnly seat can be registered from its complete role file. A genuinely absent ID needs a complete file or explicit provider/bypass choices. These Pi/false values are examples of user decisions, not defaults:
166
+ One `start AGENT` contract creates absent seats, adopts existing role files, and resumes stopped seats. No prewritten file or repeated provider/bypass decision is required for an ad-hoc start. Receipts include `session_policy`, an effective `declaration`, and the `defaulted` field list:
159
167
 
160
168
  ```sh
161
169
  team-agent start auditor --role-file /absolute/path/to/auditor.md
162
- team-agent start analyst --provider pi --model openai-codex/gpt-6-luna --bypass false --prompt "Analyze the proposed change"
170
+ team-agent start analyst --provider pi --bypass false --prompt "Analyze the proposed change"
163
171
  ```
164
172
 
165
- An imported external role goes into this team's `agents/ID.md`; the original stays untouched. A file already at that destination is used in place. Equal file/CLI definitions are accepted; conflicting provider/model/effort/bypass/prompt/profile values refuse before writes/startup. Omitted model, effort, profile and role body remain unchanged.
173
+ An imported external role goes into this team's `agents/ID.md`; the original stays untouched. A file already at that destination is used in place. Explicit CLI fields override the selected declaration. Omitted existing fields and role instructions remain unchanged; genuinely omitted defaults are identified in receipts.
166
174
 
167
- Use `agent add` when explicitly requesting strict add-only behavior. It creates and starts a new seat but refuses an occupied ID, including stopped seats. Its disaster-recovery `--force` still requires a positively dead registered old pane and no live target cohort; a live or unknown target refuses. Force is not a role-update or routine restart shortcut.
175
+ Hidden spelling aliases delegate directly to `start AGENT`. There is no second strict-add contract and no start `--force` option. Unknown ownership or liveness refuses rather than granting permission to replace a process.
168
176
 
169
177
  ## Change a registered seat
170
178
 
@@ -173,9 +181,9 @@ team-agent stop reviewer --workspace . --team project-team
173
181
  team-agent start reviewer --effort high --bypass false --workspace . --team project-team
174
182
  ```
175
183
 
176
- Stop first before applying edited instructions or launch settings. A running seat with requested changes or known drift refuses rather than claiming a successful no-op. `--prompt` replaces permanent role instructions, not a one-time task; send one-time work with `send`. `--bypass` requires explicit true/false. Matching provider is allowed, but changing engines, even by file edits, is refused; use a new ID.
184
+ Stop first before applying edited instructions or launch settings. A running seat with requested changes or known drift refuses rather than claiming a successful no-op. `--prompt` replaces permanent role instructions, not a one-time task; send one-time work with `send`. `--bypass` requires explicit true/false. A stopped seat may change provider/model with `start AGENT --provider PROVIDER --model MODEL --session fresh`. Its logical ID and history are retained. Provider changes under preserve or auto refuse and instruct fresh. A live seat with identical configuration is a no-op; any changed configuration or fresh-session request requires `stop AGENT` first.
177
185
 
178
- The original role/spec transaction and provider admission remain authoritative. A failure is not an applied update; never assume a partial/unknown native replacement was rolled back or is safe to retry. Inspect the exact scoped receipt, status and doctor first.
186
+ The original role/spec transaction and provider admission remain authoritative. A failure can follow a committed declaration or uncertain launch; never assume a partial/unknown native replacement was rolled back or is safe to retry. Inspect the exact scoped receipt, status and doctor first.
179
187
 
180
188
  ## Restart and session policy
181
189
 
@@ -187,9 +195,11 @@ team-agent agent restart reviewer --session preserve --workspace . --team projec
187
195
  team-agent team restart --session preserve --workspace . --team project-team
188
196
  ```
189
197
 
190
- New agent/team restart defaults to **auto**: proven resume when possible, otherwise only the provider's existing legal fresh path. Auto is not a guarantee of saved memory, reset permission or a fallback around native capability refusal. Preserve requests strict continuation. The old bare/explicit-path restart bridge retains its original preserve default and output.
198
+ New agent/team restart defaults to **auto**: proven resume when possible, otherwise only the provider's existing legal fresh path. Auto is not a guarantee of saved memory, reset permission or a fallback around native capability refusal. Preserve requests strict continuation. Start defaults to preserve for existing resumable context. Contextless seats and legacy providers without resume support start fresh and report `op=fresh`; launch timestamps and pending IDs alone do not require continuation. `start AGENT --session fresh` explicitly creates a new native session while retaining old history. Fresh and `--allow-fresh` cannot be combined. The old bare/explicit-path restart bridge retains its original preserve default and output.
191
199
 
192
- Single-agent restart checks scope, role/session backing and capabilities before stop; only confirmed stop allows start. On partial failure, report actual completed steps and inspect; do not blindly reset or replay. Team restart does not reactivate explicitly stopped or paused seats. Whole-team start intentionally activates the selected non-paused roster; neither command silently unpauses.
200
+ Single-agent restart checks scope, role/session backing and capabilities before stop; only confirmed stop allows start. On partial failure, report actual completed steps and inspect; do not blindly reset or replay. Team restart does not reactivate explicitly stopped or paused seats. Whole-team start also preserves explicit stops; use `start AGENT` for an intentional wake-up. Neither command silently unpauses.
201
+
202
+ Ordinary framed managed sends after a cold launch wait up to 60 seconds for initialization evidence belonging to that launch before injecting input. A readiness timeout leaves the message stored but injects nothing; inspect status/doctor before retrying. Mailbox-only delivery does not wait. Completed startup evidence survives helper exit or binary upgrades: this is not an ongoing MCP-health gate, nor proof of business acceptance.
193
203
 
194
204
  ## Explicit agent operations
195
205
 
@@ -213,12 +223,36 @@ team-agent profile init my-api --auth-mode compatible_api --workspace .
213
223
 
214
224
  A role may explicitly reference `auth_mode: compatible_api` and `profile: my-api`, with the user's chosen provider/bypass. Profile model applies only when the role has no explicit override. Subscription and custom-API capabilities differ; follow actual provider admission, not a generic promise.
215
225
 
226
+ ## Native REPL input: `--bare`
227
+
228
+ `--bare` removes message framing, not addressing safeguards. It is only for a trusted leader/operator sending one live, roster-bound non-leader worker a native input line. It is not exposed by MCP. Ordinary tasks must continue to use framed send.
229
+
230
+ ```sh
231
+ team-agent send worker '/compact' --bare --workspace . --team project-team
232
+ team-agent send --targets worker '/compact' --bare --workspace . --team project-team
233
+ team-agent send worker --bare --workspace . --team project-team -- '/compact retain unresolved issues'
234
+ ```
235
+
236
+ Only Claude, Codex and Pi are supported, including their existing canonical aliases. The content must be nonblank, one line, at most 512 UTF-8 bytes, without control characters or Unicode line separators. Admitted bytes, including surrounding spaces and a trailing semicolon, are preserved. The `--` terminator makes every following token payload, even `--socket`, `--json` or `--bare`. No `--raw` alias or message-file source is added.
237
+
238
+ Bare refuses mailbox delivery, leader targets, fanout, unmanaged `--socket`/`--pane` addressing, `--dialect` overrides, and `--task`, `--presentation-sink`, `--message-class`, `--case-id`. Use the existing comma-based `--targets worker` syntax, not a JSON array. The lifecycle commands `/exit` and `/quit` are refused; use `team-agent stop <agent>` instead.
239
+
240
+ Physical delivery requires an empty, idle native composer and verified target ownership. A busy/draft/modal/unknown surface is refused without clearing user input. Claude and Codex use unbracketed `paste-buffer -r`; Pi uses `-p -r` to avoid slash autocomplete consuming Enter. The original payload has no heading, token or Codex newline trailer; after a 100 ms floor, the owned live composer is checked and exactly one native Enter is sent. Never treat shell metacharacters as shell code.
241
+
242
+ The existing inbox JSONL remains the audit source: original content, sender, target and submission evidence are retained with `framing: "bare"`. No `events.jsonl` is created. Explicit message ids retain duplicate protection; changing framing under the same id yields `request_id_conflict`. Anonymous sends remain separate invocations. After a post-paste uncertainty, do not resend or clear the composer automatically.
243
+
244
+ `accepted: null` remains authoritative: physical submission is not proof of native-command completion. Bare does not update `first_send_at`, task attribution or session identity. A success receipt reports `session_binding: "not_reobserved"`; a session-changing command such as `/clear` may leave the previous observed session id stale. Use lifecycle-managed reset/start for state-tracked session changes; bare does not guarantee later resume continuity.
245
+
216
246
  ## Observe and finish
217
247
 
248
+ **Do not poll.** Never run `inbox`, `status` or `doctor` repeatedly, in sleep/retry loops, or as repeated tool calls. Live messages are pushed directly to terminals. Finish the current work, stop issuing tool calls, and wait for a new incoming message. Run a deliberate inspection again only after a new event, relevant state-changing action, or explicit user request; an empty inbox or busy status is not a reason. Mailbox-only messages require deliberate history review.
249
+
250
+ These three commands append one `[Advisory: do not poll]` block to human output, on the existing output stream (including stderr for human errors). With `--json`, the same guidance is the root string `polling_advisory`: one valid JSON document, never trailing non-JSON text. Existing fields, nested boolean `advisory` values and exit codes retain their meanings; strict schemas may need to allow the additional root key.
251
+
218
252
  `status` / `team status` expose the concise node projection; `--detail` does not restore historical internal dumps. Use `doctor` for diagnosis. Check actual readiness and returned recovery actions; unknown evidence is not a stopped/live proof. Stop with `stop --workspace PATH --team KEY --json` and verify scoped residual sessions, processes and owned files. Protect parent/sibling/shared resources.
219
253
 
220
254
  Workers have only `send_message`, `report_result` and optional `get_team_status`. Report completion through `report_result`; do not promote a worker to lifecycle operator. See [nested teams](https://github.com/Florious95/team-agent/blob/main/skills/team-agent/references/team-in-team.md) and [recovery](https://github.com/Florious95/team-agent/blob/main/skills/team-agent/references/recovery-runbook.md).
221
255
 
222
256
  ## Legacy command bridges
223
257
 
224
- Existing scripts may retain `quick-start`, `shutdown`, and `add-agent/start-agent/stop-agent/remove-agent/reset-agent/clone-agent/fork-agent`. These are compatibility interfaces, not the daily workflow. Use the short/team/agent forms above for new instructions. Do not mechanically rename scripts: legacy start refuses a live seat, new start can prove a no-op; bare/explicit-path legacy restart preserves its default. Old normal command output and failure evidence remain authoritative.
258
+ Existing scripts may retain hidden compatibility spellings, but new instructions use the canonical start/stop/team/agent forms above. The start aliases share one resolver, defaults, session policy, no-op/refusal rules and receipts. Bare/explicit-path legacy restart preserves its separate documented default. Startup evidence remains authoritative.
@@ -29,4 +29,4 @@ Pending.
29
29
 
30
30
  ## Next Step
31
31
 
32
- - Review role files and permissions, then run `team-agent quick-start`; use `team-agent doctor` to diagnose the team.
32
+ - Review role files and permissions, then run `team-agent start --from <team-dir>`; use `team-agent doctor` to diagnose the team.