42 Commits

Author SHA1 Message Date
Lilleman c062f8793f Kill the two surviving mutants of the headless claude calls
CI / tests (pull_request) Has been cancelled
2026-11-20 17:27:58 +01:00
Lilleman 700ac3e2ce Load no MCP server in a headless claude call that is denied tools
CI / tests (pull_request) Successful in 3m45s
2026-11-20 17:15:42 +01:00
Lilleman 5e54057f9e Answer stability round 2: 0 🔴, 2 🟡 (0 misread); report each fact of the shortfall in a subTest, and check that a transcript longer than the floor is shown whole where the records leave room
CI / tests (pull_request) Successful in 2m51s
2026-11-20 01:41:21 +01:00
Lilleman ee44c99f28 Test that the judge's failure names a room of 0 where the records overflow its prompt
CI / tests (pull_request) Successful in 3m41s
2026-11-20 01:36:22 +01:00
Lilleman e062f71d47 Answer stability round 1: 0 🔴, 2 🟡 (0 misread); tie the help's transcript floor to TRANSCRIPT_FLOOR, split the floor test, and name the shortfall in the judge's failure, with the prose round's wording 2026-11-20 01:24:21 +01:00
Lilleman 32a81b0257 Test that a head leaving exactly the floor still shows the transcript
CI / tests (pull_request) Successful in 3m7s
2026-11-20 00:59:11 +01:00
Lilleman f3af5e23fd Stop the judge rather than show it under 50,000 characters of the transcript, read a twice-named record once, test the forge timeout and decide the merge stop from the pull requests found 2026-11-20 00:45:49 +01:00
Lilleman 1845d8ff37 Answer prose round 1: 0 🔴, 4 🟡 (0 misread); say that the task and the transcript are searched, and where a quote may not run across them 2026-11-19 22:36:14 +01:00
Lilleman d954fb0026 Answer security round 1: 0 🔴, 0 🟡 (0 misread); keep a judge's evidence quote from running across the task and the tail, and name the task in the schema and the rejection message 2026-11-19 22:24:24 +01:00
Lilleman 14123912d9 Let the judge quote the worker's task as evidence
CI / tests (pull_request) Has been cancelled
2026-11-19 21:21:54 +01:00
lilleman 2806a8edc9 Kill the mutants that survived on the lines the release review changed 2026-11-19 13:44:12 +01:00
lilleman cb3135bc8c Answer principles round 1: 1 🔴, 7 🟡 (0 misread); every effect is handed in as an argument, no test patches a global, and the traces principle scopes tools that run on their user's machine
CI / tests (pull_request) Successful in 3m43s
2026-11-19 12:42:03 +01:00
lilleman 9bc2887ee8 Answer architecture round 3: 0 🔴, 1 🟡 (0 misread); every suite sits beside the module it tests, and test_loop_claude.py holds only loop-claude's own
CI / tests (pull_request) Successful in 5m17s
2026-11-19 11:43:38 +01:00
lilleman df210aaeca Answer architecture round 2: 0 🔴, 3 🟡 (0 misread); the cycle's and the chunk run's suites sit beside their modules, and the claim rule and the model override have one home each
CI / tests (pull_request) Successful in 3m23s
2026-11-19 11:29:32 +01:00
lilleman b8595db647 Answer architecture round 1: 0 🔴, 36 🟡 (0 misread); chunk_driver holds what both drivers share, and Jira and review-state reads have one home each
CI / tests (pull_request) Successful in 5m11s
2026-11-19 11:14:44 +01:00
lilleman d06eea3149 Kill the mutants of the judged shells, the named process's lookup and a tool named as a third word
CI / tests (pull_request) Has been cancelled
2026-11-19 10:11:51 +01:00
lilleman a49a790a36 Answer product-owner round 2: 0 🔴, 1 🟡 (0 misread); judge a shell again once it used CPU and went idle, retry a failed judgment, and name a process without a command line 2026-11-19 10:02:12 +01:00
lilleman f3af8fda13 Answer product-owner round 1: 0 🔴, 3 🟡 (0 misread); check a shell that holds a heavy slot, name the process's command, and judge each shell once 2026-11-19 09:56:43 +01:00
lilleman dbfe87d36f Sort the stuck check's imports by their first binding 2026-11-19 09:26:57 +01:00
lilleman 53ef45dbd8 Sort the stuck check's imports by their first binding 2026-11-19 09:26:30 +01:00
lilleman d251d50472 Kill the mutants of the evidence separator and the CPU sum 2026-11-19 09:16:14 +01:00
lilleman a12710ebc4 Kill the mutants of the stuck check, its rung and the /proc readings it hands the investigation 2026-11-19 09:06:26 +01:00
lilleman 08038d93eb Judge a worker that stalls before its result with a headless claude, and report the verdict without a kill
CI / tests (pull_request) Failing after 4m3s
2026-11-19 08:55:14 +01:00
lilleman 1e19f7759d Answer stability round 2: 1 🔴, 1 🟡 (0 misread); say a subagent's shells are stopped like the worker's own, and share one polling helper in the tests
CI / tests (pull_request) Has been cancelled
2026-11-19 08:07:00 +01:00
lilleman 46dc401736 Merge remote-tracking branch 'origin/main' into item-420
# Conflicts:
#	bin/loop_claude/test_loop_claude.py
#	bin/loop_effects/test_effects.py
2026-11-19 07:46:23 +01:00
lilleman e55b8826fb Merge remote-tracking branch 'origin/main' into item-310
# Conflicts:
#	bin/agent_cycle/interrupt.py
#	bin/agent_cycle/run.py
#	bin/agent_cycle/states.py
#	docs/reading-cost.md
#	todo.md
2026-11-19 04:47:53 +01:00
lilleman 0a77e2af00 Merge origin/main into item-414
CI / tests (pull_request) Successful in 5m11s
2026-11-19 03:31:12 +01:00
lilleman d652ab20d3 Answer product-owner round 1: 1 🔴, 1 🟡 (1 misread); DM through the checkout of a held-up issue's repo and show the DM in the window 2026-11-19 02:44:22 +01:00
lilleman ed58379da0 Answer stability round 1: 0 🔴, 3 🟡 (0 misread); give an awaits-a-slot chunk no since, print the time in UTC and keep the status lines tested once 2026-11-19 02:17:48 +01:00
lilleman 8e210eaa09 Print each chunk that asks a question and since when each parked chunk awaits its gate in do-epic --status
CI / tests (pull_request) Failing after 3m23s
2026-11-19 00:40:16 +01:00
lilleman 4ed8e1a767 Merge remote-tracking branch 'origin/main' into item-411
CI / tests (pull_request) Has been cancelled
# Conflicts:
#	README.md
#	bin/do_epic/chunk_prompt.py
#	bin/do_epic/main.py
#	bin/do_epic/test_main.py
2026-11-19 00:08:28 +01:00
lilleman d49831c387 Answer prose round 1: 1 🔴, 6 🟡 (0 misread); reword the help, comments, docstrings and the decision, and read a Gitea draft in the commits before 2026-11-18 23:55:19 +01:00
lilleman 89a7cfe65a Park on a new review once requested changes are answered, never park a Gitea draft, and count a slot only for a window that started 2026-11-18 23:53:28 +01:00
lilleman d0ecdd6359 File a finding and record a decision in a do-epic epic, and give the judge the epic's records beside the repo's
CI / tests (pull_request) Failing after 4m22s
2026-11-18 23:13:43 +01:00
lilleman 1df19f4ac8 Keep a parked chunk out of the epic's and its repo's running slots 2026-11-18 23:11:31 +01:00
lilleman 06f27feb3e Answer stability round 1: 0 🔴, 6 🟡 (1 misread); name a checkout by its real path, pick a chunk's repo once, test the step line, the repo's model and ladder, and a clone 2026-11-18 22:50:30 +01:00
lilleman 38e39c93d1 Run do-epic <EPIC-KEY>: a tmux window per chunk, under the epic's cap and each repo's parallel-chunks, never beside loop-claude 2026-11-18 20:49:12 +01:00
lilleman c4d06c55b5 Name the driver of a cycle in the text its worker, judge and human read 2026-11-18 20:34:40 +01:00
lilleman f62ea168fe Answer prose round 1: 0 🔴, 6 🟡 (0 misread); reword the docstrings, the README entries and item 401's detail 2026-11-18 17:07:03 +01:00
lilleman 885bf37891 Answer stability round 1: 0 🔴, 6 🟡 (0 misread); pin the answer destinations in the prompts, name Brief's fields, drop the duplicate import test 2026-11-18 17:07:03 +01:00
lilleman a7153c78fd Sort the imports the move reordered 2026-11-18 16:24:27 +01:00
lilleman 54fc339fb0 Move loop-claude's cycle, judge, interrupt, tmux and routing into agent_cycle
CI / tests (pull_request) Failing after 3m50s
2026-11-18 16:21:35 +01:00