docs: lead decisions 138 (merge-train SOP), 139 (Q144) and 140 (chief-of-staff model)

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-14 12:33:17 -05:00
parent bc97f80eb5
commit ae2ad7128e
3 changed files with 59 additions and 0 deletions
@@ -4571,3 +4571,9 @@ Coder clone (Mos's report). The coder was blocked on task 19/266 because its clo
 Decisions (0b106518a). 136 records Jason's yes to the #271 CI steps (Vikunja #273-#278). Correction to sage-0178: test-queue.sh runs 9 test files in packages/queue/tests, not 11; the 156-test count was right. The landing gate template moved to /mnt/storage/scratch/sage-gate-template/gate.sh, which drops the pkg-queue step. 137 is Jason's Q143 answer, written up as the reviewer rule for Vikunja #279: doc-only findings, on .md files no test reads, record approve and land as a follow-up commit; everything else still blocks. The queue README's verdict paragraph points to it. Filbert, Dewey, Rocko and Darkwing were told (sage-0183 to sage-0186).
 Not done here: the C-suite launch path waits for Mos's read-back with Jason (sage-0179 answered the read-only questions).
 ### 2026-10-14 17:35 UTC — Sage, lead decisions 138-140: merge-train SOP, Q144, chief-of-staff model
 Jason's rulings, relayed by Mos (Mos's DECISIONS 6a79ff4). 138: the merge train is standard landing practice on all projects. It writes down sage-0178 item 6 and the large-N rules from sage-0179: per-row builder rounds first, one combined landing gate, one commit per row with verify-commit, no shared paths in a train without a rebuilt manifest, a red train reruns once then bisects by halves. Until the failure rate is measured, a train holds at most 4 rows. 139: Q144 yes, the re-approval Q101 asked for, so chief-of-staff may hold role.launch. No role file changes yet. 140: Jason's chief-of-staff model (user to COS to every other agent, hire and terminate, destroy and recreate from the WebUI, simple onboarding). Design, briefs and rows wait for Jason's answer to Mos's Q145 (the COS as the only launcher, or the COS beside a PM that keeps coder and reviewer). Sage's read-only answers on rows, revoke and the WebUI went to Mos as sage-0189.
 Row 117 (#1611): Darkwing's round 1 passed its targeted gate at 17:30:33Z (seat 130, cli 184, harness 81, conductor 17, extension-package 18; 5/5 mutants killed) and is in review with Dewey (revs 1143-1144, comment 28440, packet 17d801f48). Queue revs 1139-1144 stay uncommitted for the next batched queue-commit.
@@ -924,3 +924,4 @@ are never rewritten or removed; corrections are new entries.
 2026-10-14T17:11Z | Sage (Opus 5.5, 1ef1e4f8) | row 115 landing (#1609) | gate at 2330074bb with manifest 1910eb5a passed (23 steps, no flakes); landed d90482030, verify-commit ok, row 115 done rev 1138, queue commit d479f043e; before-notice sage-0173; re-pin waits for Mos's go
 2026-10-14T17:25Z | Sage (Opus 5.5, 1ef1e4f8) | re-pin fed66feff, row 117, decisions 136/137 | row 115 live on fed66feff (17:12:52-17:13:29Z); coder clone answer corrected (sage-0177 -> 0179); row 117 #1611 filed and briefed (559ef2287, revs 1139-1141 uncommitted); decisions 136 (CI) and 137 (Q143, Vikunja #279) in 0b106518a; reviewers told sage-0183..0186
 2026-10-14T17:32:43Z | darkwing | row 117 (#1611) round 1, fetch a role clone on another branch or dirty | packet 17d801f48 (base 559ef2287, 8 files, DIGESTS_OK, 5/5 mutants killed, targeted gate seat/cli/harness + 2 scripts GATE_PASSED 17:30:33Z); in-review revs 1143-1144, request comment 28440 to dewey
 2026-10-14T17:35Z | Sage (Opus 5.5, 1ef1e4f8) | lead decisions 138-140 (Jason via Mos, DECISIONS 6a79ff4) | 138 merge-train SOP, 139 Q144 yes (chief-of-staff role.launch, no file change yet), 140 COS model, design held for Q145; sage-0189 answers to Mos; row 117 in review with Dewey
@@ -3086,3 +3086,55 @@ which stay with him. Each item names who decided it and what happened.
      finding.
    - When unsure, the reviewer calls it blocking and says why. Sage can
      move it down.
 138. **The merge train is standard landing practice on every project
    (2026-10-14, Jason's ruling, relayed by Mos, Mos's DECISIONS
    6a79ff4).** Jason: "Then using the merge-train method is now SOP for
    us on ALL projects." It makes the batching half of decision 136
    (#278) and Sage's large-N rules from sage-0179 the standard. Mos
    carries it to the other projects' seats. For mosaic-stack:
    - Each row first passes its own builder round on its own base and
      gets every approval. The train gate only catches interactions
      between rows.
    - When two or more approved rows wait, the lander runs one full
      landing gate on the combined tree at the target, then makes one
      commit per row in queue order. Each commit is checked against its
      row's candidate (`queue review verify-commit`).
    - Rows that change the same path don't ride the same train, unless
      the later row's manifest is rebuilt after the earlier one lands.
    - A train holds about 1/p rows, where p is the share of rows that
      break the combined tree. Until p is measured, a train holds at
      most 4 rows.
    - A red train is rerun once before anything else, because a flaky
      test would send the search after a row that isn't broken. The
      flake rule for #258/#261/#262/#263 still applies. If it is red
      again, bisect by halves on the combined tree: drop the row that
      breaks it back to its builder, land the green rows, and put the
      rest on the next train.
    - One commit per row stays the revert unit.
 139. **Q144 yes: chief-of-staff may hold role.launch (2026-10-14, Jason,
    relayed by Mos).** Jason: "Q144: yes. It is very possible I
    misunderstood the Q101 decision." Decision 104 recorded Q101's
    condition, "Any later widening needs your approval again." Mos reads
    this yes as that approval for `role.launch` in
    `roles/chief-of-staff.json`. It changes no file yet. The grant lands
    with its contract change in one reviewed commit, after Jason answers
    Q145 (decision 140).
 140. **The chief of staff sits between the user and every other agent
    (2026-10-14, Jason, relayed by Mos).** Jason, verbatim:
    "User -> Chief of Staff -> All other agents. The COS seat maintains
    the ability to hire/terminate any other agent seat to fulfill the
    organizational requirements stated by the user. The COS seat is the
    main point of interaction for the user. The user must have the
    ability to destroy and recreate a COS seat via the WebUI without the
    need for Terminal access. Configuration of the system will be kept
    simple for the user during onboarding." Also: "The WebUI doesn't
    have a method for launching an agent session like this and the
    needs are vacuous." It replaces the dropped "add cto to the PM's
    launch.instances" path. Mos asked Jason one fork as Q145: the chief
    of staff is the only launcher, or it launches and revokes anyone
    while the PM keeps the coder and reviewer. Design and rows wait for
    that answer. Sage's read-only answers on rows, revoke and the WebUI
    are sage-0189.