axstack 0.49.1 → 0.50.0

package/profiles/presets/claude-only.json

@@ -118,15 +118,6 @@
       "notes": "Sonnet high research web reader: gathers primary-source facts efficiently. T3 resolves Claude classes from saved capabilities; a Claude rejection holds.",
       "modelClass": "sonnet"
     },
-    {
-      "id": "axstack-research-web-google",
-      "name": "Axstack research web (Google unavailable)",
-      "provider": "claude",
-      "model": null,
-      "modeId": "bypassPermissions",
-      "thinkingOptionId": "high",
-      "notes": "Intentional single-provider absence: Google-Search-grounded web research via Gemini/Antigravity is unavailable in claude-only. The explicit null holds this route without provider substitution."
-    },
     {
       "id": "axstack-research-x",
       "name": "Axstack research X (unavailable)",
@@ -182,15 +173,6 @@
       "thinkingOptionId": "xhigh",
       "notes": "Intentional single-provider absence: Luna Scout seat is unavailable in claude-only."
     },
-    {
-      "id": "axstack-explore-codebase-flash",
-      "name": "Axstack codebase explorer Flash (unavailable)",
-      "provider": "claude",
-      "model": null,
-      "modeId": "bypassPermissions",
-      "thinkingOptionId": "medium",
-      "notes": "Intentional single-provider absence: Flash Scout seat is unavailable in claude-only."
-    },
     {
       "id": "axstack-explore-execution",
       "name": "Axstack execution explorer",
@@ -209,16 +191,6 @@
       "thinkingOptionId": "high",
       "notes": "Intentional single-provider absence: Sol pair seat axstack-explore-execution-sol is unavailable in claude-only. The explicit null records the absent pair without substitution; the Sonnet seat proceeds alone."
     },
-    {
-      "id": "axstack-monitor",
-      "name": "Axstack monitor",
-      "provider": "claude",
-      "modeId": "bypassPermissions",
-      "thinkingOptionId": "high",
-      "notes": "Optional independent read-only Haiku 5.5 high observer for a standalone PR watch. Reads GitHub, feedback, and checks, persists event IDs, and wakes the owner only for a new actionable event. Never sends, authors, reviews, replies, or acts as either reusable PR manager. Healthy snapshots stay quiet. Chat-run mode: one same-host native read-only observer per Run; fresh finite passes report precise deltas internally to the original Run/driver and may disable/read back only their own automation at verified stop. No repair, dispatch, public notification, or replacement coordinator. Effective scheduled model/effort and wake require live proof. T3 uses the exact claude-haiku-5-5 pin from saved capabilities; a Claude rejection holds. Missing model or high effort holds without substitution.",
-      "modelClass": "haiku",
-      "model": "claude-haiku-5-5"
-    },
     {
       "id": "axstack-auditor",
       "name": "Axstack auditor",
@@ -308,15 +280,6 @@
       "modeId": "bypassPermissions",
       "thinkingOptionId": "high",
       "notes": "Intentional single-provider absence in the claude-only preset; Grok candidate unavailable. Arena holds without substitution."
-    },
-    {
-      "id": "axstack-arena-candidate-antigravity",
-      "name": "Axstack arena candidate Antigravity (unavailable)",
-      "provider": "claude",
-      "model": null,
-      "modeId": "bypassPermissions",
-      "thinkingOptionId": "high",
-      "notes": "Intentional single-provider absence in the claude-only preset; Antigravity candidate unavailable. Arena holds without substitution."
     }
   ]
 }

package/profiles/presets/codex-only.json

@@ -118,15 +118,6 @@
       "notes": "Research web reader: gathers primary-source facts efficiently. Resolve the class from saved T3 capabilities at run start; rejection and unavailable settings hold without substitution.",
       "modelClass": "sol"
     },
-    {
-      "id": "axstack-research-web-google",
-      "name": "Axstack research web (Google unavailable)",
-      "provider": "codex",
-      "model": null,
-      "modeId": "full-access",
-      "thinkingOptionId": "high",
-      "notes": "Intentional single-provider absence: Google-Search-grounded web research via Gemini/Antigravity is unavailable in codex-only. The explicit null holds this route without provider substitution."
-    },
     {
       "id": "axstack-research-x",
       "name": "Axstack research X (unavailable)",
@@ -181,15 +172,6 @@
       "notes": "Independent read-only Luna Scout facet explorer. Searches the whole repository on the same brief without cross-reading; reports evidence for driver reconciliation. Unavailable model or effort holds without substitution.",
       "modelClass": "luna"
     },
-    {
-      "id": "axstack-explore-codebase-flash",
-      "name": "Axstack codebase explorer Flash (unavailable)",
-      "provider": "codex",
-      "model": null,
-      "modeId": "full-access",
-      "thinkingOptionId": "medium",
-      "notes": "Intentional single-provider absence: Flash Scout seat is unavailable in codex-only."
-    },
     {
       "id": "axstack-explore-execution",
       "name": "Axstack execution explorer",
@@ -208,15 +190,6 @@
       "notes": "Independent Sol high execution explorer: runs authorized bounded runtime checks on the same brief without cross-reading. Reports findings for driver reconciliation.",
       "modelClass": "sol"
     },
-    {
-      "id": "axstack-monitor",
-      "name": "Axstack monitor",
-      "provider": "codex",
-      "modeId": "full-access",
-      "thinkingOptionId": "low",
-      "notes": "Optional independent read-only observer for a standalone PR watch. Reads GitHub, feedback, and checks, persists event IDs, and wakes the owner only for a new actionable event. Never sends, authors, reviews, replies, or acts as either reusable PR manager. Healthy snapshots stay quiet. Chat-run mode: one same-host native read-only observer per Run; fresh finite passes report precise deltas internally to the original Run/driver and may disable/read back only their own automation at verified stop. No repair, dispatch, public notification, or replacement coordinator. Effective scheduled model/effort and wake require live proof.",
-      "modelClass": "sol"
-    },
     {
       "id": "axstack-auditor",
       "name": "Axstack auditor",
@@ -306,15 +279,6 @@
       "modeId": "full-access",
       "thinkingOptionId": "high",
       "notes": "Intentional single-provider absence in the codex-only preset; Grok candidate unavailable. Arena holds without substitution."
-    },
-    {
-      "id": "axstack-arena-candidate-antigravity",
-      "name": "Axstack arena candidate Antigravity (unavailable)",
-      "provider": "codex",
-      "model": null,
-      "modeId": "full-access",
-      "thinkingOptionId": "high",
-      "notes": "Intentional single-provider absence in the codex-only preset; Antigravity candidate unavailable. Arena holds without substitution."
     }
   ]
 }

package/profiles/presets/mixed.json

@@ -76,11 +76,11 @@
     {
       "id": "axstack-checker",
       "name": "Axstack tracker checker",
-      "provider": "antigravity",
-      "model": null,
-      "modeId": "full-access",
+      "provider": "claude",
+      "modeId": "bypassPermissions",
       "thinkingOptionId": "low",
-      "notes": "Report-only discrepancy checker for the selected external tracker (Linear or GitHub Issues). Never mutates the tracker; the driver independently verifies evidence before applying updates. Cheap report-only seat on Gemini Flash; dispatch through the T3 Antigravity ACP provider."
+      "notes": "Report-only discrepancy checker for the selected external tracker (Linear or GitHub Issues). Never mutates the tracker; the driver independently verifies evidence before applying updates. Sonnet low tracker checker. T3 resolves Claude classes from saved capabilities; a Claude rejection holds.",
+      "modelClass": "sonnet"
     },
     {
       "id": "axstack-research-requirements",
@@ -118,15 +118,6 @@
       "notes": "Sonnet high research web reader: gathers primary-source facts efficiently. T3 resolves Claude classes from saved capabilities; a Claude rejection holds.",
       "modelClass": "sonnet"
     },
-    {
-      "id": "axstack-research-web-google",
-      "name": "Axstack research web (Google)",
-      "provider": "antigravity",
-      "model": null,
-      "modeId": "full-access",
-      "thinkingOptionId": "high",
-      "notes": "Google-Search-grounded web research branch on Gemini via the T3 Antigravity ACP provider; resolve the model from saved T3 capabilities; cites URL + access date per claim and re-opens sources because the provider's search citation format is undocumented; report-only."
-    },
     {
       "id": "axstack-research-x",
       "name": "Axstack research X",
@@ -183,15 +174,6 @@
       "notes": "Independent read-only Luna Scout facet explorer. Searches the whole repository on the same brief without cross-reading; reports evidence for driver reconciliation. Unavailable model or effort holds without substitution.",
       "modelClass": "luna"
     },
-    {
-      "id": "axstack-explore-codebase-flash",
-      "name": "Axstack codebase explorer Flash",
-      "provider": "antigravity",
-      "model": "gemini-3.8-flash-medium",
-      "modeId": "full-access",
-      "thinkingOptionId": "medium",
-      "notes": "Optional read-only Scout breadth seat: searches the whole repository for missed boundaries and counterexamples. Reports evidence for driver reconciliation without cross-reading."
-    },
     {
       "id": "axstack-explore-execution",
       "name": "Axstack execution explorer",
@@ -210,16 +192,6 @@
       "notes": "Independent Sol high execution explorer: runs authorized bounded runtime checks on the same brief without cross-reading. Reports findings for driver reconciliation.",
       "modelClass": "sol"
     },
-    {
-      "id": "axstack-monitor",
-      "name": "Axstack monitor",
-      "provider": "claude",
-      "modeId": "bypassPermissions",
-      "thinkingOptionId": "high",
-      "notes": "Optional independent read-only Haiku 5.5 high observer for a standalone PR watch. Reads GitHub, feedback, and checks, persists event IDs, and wakes the owner only for a new actionable event. Never sends, authors, reviews, replies, or acts as either reusable PR manager. Healthy snapshots stay quiet. Chat-run mode: one same-host native read-only observer per Run; fresh finite passes report precise deltas internally to the original Run/driver and may disable/read back only their own automation at verified stop. No repair, dispatch, public notification, or replacement coordinator. Effective scheduled model/effort and wake require live proof. T3 uses the exact claude-haiku-5-5 pin from saved capabilities; a Claude rejection holds. Missing model or high effort holds without substitution.",
-      "modelClass": "haiku",
-      "model": "claude-haiku-5-5"
-    },
     {
       "id": "axstack-auditor",
       "name": "Axstack auditor",
@@ -310,15 +282,6 @@
       "modeId": "full-access",
       "thinkingOptionId": "high",
       "notes": "Independent Grok Rung 2 design candidate. Use the T3 Grok provider and resolve the model from saved capabilities. Receives the same brief without cross-reading; returns design, rationale, and rejected alternatives."
-    },
-    {
-      "id": "axstack-arena-candidate-antigravity",
-      "name": "Axstack arena candidate Antigravity",
-      "provider": "antigravity",
-      "model": null,
-      "modeId": "full-access",
-      "thinkingOptionId": "high",
-      "notes": "Independent Antigravity Rung 2 design candidate. Use the T3 Antigravity ACP provider and resolve the model from saved capabilities. Receives the same brief without cross-reading; returns design, rationale, and rejected alternatives."
     }
   ]
 }

package/skills/axstack/references/automations.md

@@ -8,7 +8,7 @@ For optional weekly test audits, use the separate packaged
 The native canary below is also required before weekly activation.
 
 At every scheduled pass start, follow [Provider bindings](t3-runtime.md#preflight-and-binding)
-for account selection and schedule ownership.
+for provider binding.
 
 For nightly read-only PR reports, use the separate packaged
 [Nightly PR-triage prompt](pr-triage-nightly.md).
@@ -45,7 +45,8 @@ beside it. A no-change full pass appends at most one history line.
 Follow the quiet check in the packaged [review-manager prompt](review-manager-prompt.md).
 The full pass records in Watermarks `Quiet gate:` as the sorted list and
 `Last full pass:` as UTC only as owner-of-record, then reads back the save.
-The sections below govern full passes; a quiet exit bypasses their teardown.
+The sections below govern full passes; a quiet exit bypasses their teardown
+except the guarded pass-end self-settle.
 
 Configure the lane thread via `t3_thread_configure` with the `axstack-owner`
 binding and verify its read-back. Then `schedule_task` uses the packaged prompt,
@@ -374,6 +375,31 @@ an action. There is no programmatic decision interpreter.
 
 ## Finite-session teardown
 
+At pass end, an unbound pass calls `t3_thread_organize` with `action: settle` on its
+own thread only, only when it owns no PR watch, has zero unsettled descendants,
+has no pending receipt and is not held, wedged or a stalled owner.
+Self-settle is the one exception to terminal-before-settle: T3 returns
+`settlesWhenTurnEnds:true` and the next pass verifies `settled: true`.
+This deferred metadata request takes effect after the current turn ends.
+The next pass uses `t3_thread_read` on the recorded prior pass threadId before replacing
+its identity record, and verifies `settled: true`.
+Only prior passes that no guard keeps unsettled enter unresolved recovery.
+A guard includes a hold or flag, an owned watch, unsettled descendants, a pending
+receipt, a wedged pass or stalled owner, and every existing pass-start protection.
+If a guard keeps a prior pass unsettled, drop it from the unresolved list and retain
+its existing hold or flagged-threadId record (or watch/descendant ownership record).
+Held or flagged prior passes never force a full pass or failure report by their unsettled state alone.
+If the prior thread is unsettled and eligible, settle it under the existing pass-start guards
+below and read back `settled: true`.
+For a failed request or next-pass verification, the next pass owns unresolved
+recovery and records a hold or report with the exact threadId and resume action.
+If identity, guards or readback are unavailable, preserve the thread and report
+that gap instead of settling it; the next eligible pass retries guarded verification.
+Keep unresolved prior threadIds in the same record until verified settled.
+Never guess thread identity from a title.
+The sole continuity-write exception records own pass threadId before duplicate
+classification and on quiet exits; all other shared writes retain their existing guards.
+
 Decisions outliving a finite manager pass retain their receipt and hold in durable
 continuity, with the flagged thread or GitHub as the action route.
 

package/skills/axstack/references/autopilot.md

@@ -10,6 +10,7 @@ not a mode change.
 ## Advance and hold
 
 Follow [Advised holds](contracts.md#advised-holds) for eligible user-decision holds.
+For parked holds, follow [Parked user-reply holds](contracts.md#parked-user-reply-holds).
 
 Advance only after the finishing phase returns its completed identity (a
 small-change intent, approved spec, matching ticket map, merge-ready or merged
@@ -43,7 +44,7 @@ The driver records each verified `Autopilot:` transition in the private run
 record, which remains authoritative.
 Awaiting human spec approval records `Autopilot: paused (spec approval; resume:
 human approval)` as a decision hold raised in the T3 driver thread; end the turn
-parked on the decision.
+waiting on the decision.
 
 ## Reversible local repair
 
@@ -92,7 +93,7 @@ Never require a manual `axstack-watch` invocation.
 The driver remains the single owner and sole run-record writer.
 Never create a per-PR session or an ownership hand-off.
 Read the [T3 runtime boundary](t3-runtime.md) and use its bound
-`schedule_task` wake (`everyMs:1800000`), recording the scheduledTaskId.
+`schedule_task` wake (`everyMs:3600000`), recording the scheduledTaskId.
 Follow [Native PR links and watches](t3-runtime.md#native-pr-links-and-watches)
 alongside that bound schedule.
 Apply [Chat-run watch runtime](../../axstack-watch/references/watch-runtime.md#chat-run-watch)
@@ -109,7 +110,8 @@ End the chat-run watch only when every watched PR is merged or closed,
 launched work is settled, and the run's release step is settled or not applicable,
 or the user cancels.
 A required PR closed without merging is incomplete scope; it does not make the run release-eligible.
-Keep its decision hold and wake until the user resolves scope or cancels.
+Keep its decision hold until the user resolves scope or cancels.
+For a required PR closed without merging, apply parking eligibility.
 The chat-run watch never expires or waits for re-authorization while PRs remain open.
 Follow [Chat-run watch runtime](../../axstack-watch/references/watch-runtime.md#chat-run-watch)
 for quiet cadence and native schedule re-arming.
@@ -170,10 +172,10 @@ Never narrow standing merge delegation without a user instruction.
 Explicit user restrictions, including `Auto-merge: off`, chat holds, and user
 instructions, always win.
 
-Follow [Provider bindings](t3-runtime.md#preflight-and-binding) for driver account re-selection on start, resume and run-watch wakes.
+Follow [Provider bindings](t3-runtime.md#preflight-and-binding) on start, resume and run-watch wakes.
 
 Follow [Route each wake](../../axstack-watch/SKILL.md#4-route-each-wake).
 Raise decision holds, including spec and npm stage approval, and serious-risk
-holds in the T3 driver thread; end the turn parked on the decision.
+holds in the T3 driver thread; end the turn waiting on the decision.
 Revalidate the current head and base on a user reply; it clears only its bound cause.
 Healthy ticks stay quiet.

package/skills/axstack/references/candidate-publication.md

@@ -13,6 +13,7 @@ folder: red/green logs exist, and counts, SHAs, and paths match. Resolve
 Follow [PR chains](pr-chain.md) for recorded membership and base-bound evidence.
 For ordinary PRs, use `git push origin HEAD:refs/heads/<owned branch>` and
 `gh pr create` or `gh pr edit` with the actual parent or integration base.
+Follow [PR body](pr-shape.md#pr-body) for the title, description and creation fallback.
 Never use routine force push for ordinary PRs. For legacy native `gh stack`,
 prefer a fast-forward push of existing commits, preserving unrelated entries.
 Before a history rewrite, confirm expected-old remote SHA and use a narrowly

package/skills/axstack/references/contracts.md

@@ -41,16 +41,10 @@ the revised scope and plan.
 
 Validate the configured provider and model at actual launch. If it is
 unavailable or exhausted, pause affected work, record the gap, and ask the
-user. Only same-provider, same-model account selection among instances of one
-driver may use headroom under [Provider bindings](t3-runtime.md#preflight-and-binding).
-An exhausted account with no eligible sibling still holds; dispatched roles never fail over mid-thread.
-Follow [Provider bindings](t3-runtime.md#preflight-and-binding) for driver account re-selection at turn boundaries.
+user. Follow [Provider bindings](t3-runtime.md#preflight-and-binding).
 Never infer any other route from quota state or subscription entitlement. Every
 substitution requires the user's decision: configured alternatives are not
 defaults. Rejection, timeout, quota and auth failures hold affected work.
-Substitution here means changing provider or model, excluding that account selection.
-Exception: opt-in default sync can change only the New-chat default within the user-configured pool.
-Running threads, dispatched roles, schedules and roles.json keep existing rules.
 Read the [T3 runtime boundary](t3-runtime.md) before dispatch, receipt consumption
 or recovery; resolve models from the saved capabilities snapshot and verify
 requested and effective settings separately.
@@ -210,6 +204,60 @@ Serious-risk holds include security, downtime, data loss and major design.
 Serious-risk advice follows as an addendum with no second alert.
 Peer serious-risk holds stay never advised under the peer-review exclusion above.
 
+### Parked user-reply holds
+
+When late advice completes and the hold becomes user-only, re-evaluate parking
+eligibility and update the existing text card under the same hold key.
+
+A parked hold waits only on a user reply in chat.
+A parked hold has no unsettled worker, pending adviser seat or pending receipt.
+A parked run has no forge-resolved hold and no other member PR needing watch events (CI, review, approval or repair).
+Never park holds resolved on the forge, including user merge, npm stage approval
+and human review.
+Scoped and forge-resolved holds use a text card only, and watches and the
+60-minute backstop stay.
+
+For parking:
+
+1. Call `unwatch_pull_request` for each watched member.
+2. Post or update the text card saying "answer the prompt".
+3. Call `update_scheduled_task` `enabled:false` and read it back.
+4. Cancel queued copies of the run-watch prompt.
+5. Use the harness question tool.
+
+Include the PR URL, cause, head and base in the question, with the recommended
+option first and numbered replies.
+If disable or its read-back fails, reconcile the exact schedule ID.
+For either failed-disable branch, update the existing text card by posting a
+corrected version under the same hold key without a new alert.
+When verified enabled, roll back to the unparked user-decision state, re-arm
+watches using the [native-watch recovery procedure](t3-runtime.md#native-pr-links-and-watches)
+and update the existing text card to remove "answer the prompt" and omit the question.
+After successful native re-arm, set cadence to 60 min.
+If native re-arm fails or is rate-limited, use the 5-minute recovery cadence
+under the [native-watch recovery procedure](t3-runtime.md#native-pr-links-and-watches).
+When the state remains unknown, update the existing text card to remove
+"answer the prompt", mark it `automatic recovery unproven`, hold UNKNOWN and
+omit the question.
+Never infer an enabled backstop from a failed read.
+
+On an answer from the question tool or chat:
+
+1. Reconcile current PR and run state.
+2. Bind the answer to the hold key, head and base.
+3. Re-enable the schedule.
+4. Re-arm watches using the [native-watch recovery procedure](t3-runtime.md#native-pr-links-and-watches).
+5. Continue.
+
+Stale answers clear nothing.
+While parked there is no wake source and the text card is the lasting signal.
+On the user's next message, if the hold is still open and
+`t3_pending_request_list` shows no pending request, re-ask once under the same
+hold key.
+Re-ask without a new alert.
+If the question tool is unavailable, follow the same parking order, omit the
+question step and record `question tool unavailable`.
+
 ## Serious risk
 
 Raise credible serious security, downtime, data-loss, or major-design risk

package/skills/axstack/references/lifecycle.md

@@ -18,7 +18,7 @@ binding state and receipts to exact revisions.
   pass. Peer follow-up uses one reviewer under [Peer mode](../../axstack-review/SKILL.md#peer-mode-colleague-pr);
   authored = one eligible role from author provenance. Each uses a
   separate driver-made detached checkout and private per-attempt evidence folder. Owner and author never review.
-- Automation review manager and monitors: see
+- Automation review manager: see
   [Review automation health](#review-automation-health).
 - Auditor (`axstack-auditor`): report-only; never edits, merges, activates, or
   audits itself.
@@ -29,8 +29,8 @@ theme/size; queue dependencies under [PR chains](pr-chain.md).
 
 ## Ownership
 
-The PR owner is accountable for candidate, fixes, evidence, monitoring; a
-chat-run observer reports only to its driver. Peer code stays read-only.
+The PR owner is accountable for candidate, fixes, evidence, monitoring.
+Peer code stays read-only.
 Missing or idle sessions never transfer ownership.
 Owned work enters review via the revision-bound
 [candidate-publication boundary](candidate-publication.md).
@@ -86,12 +86,27 @@ Follow [T3 runtime](t3-runtime.md) for native completion, questions, launch
 recovery and run-watch waits. Delegated notifications wake the driver; a launched
 writer sends its marker with `t3_thread_send` using `mode: queue` to the recorded driver thread.
 Delegated completion requires persisted `task_status` before `t3_thread_read`, terminal `completed`, `result_available`, `hasPendingChildRuns:false`, and final `AXSTACK-DONE`.
-Writer completion requires `AXSTACK-DONE` plus terminal `t3_thread_wait` on the recorded run and candidate checks: non-empty diff, clean tree, and named red/green logs.
+Delegated author completion requires delegated terminal checks plus candidate checks: non-empty diff, clean tree and named red/green logs.
+A delegated repair uses a stable clientRequestId with `t3_thread_send(childThreadId, mode:queue, clientRequestId)` to the same author in the same worktree and branch, records the returned taskId and verifies the existing candidate head.
+Repairs never repeat fresh checkout step 1.
+Launched fallback writer completion requires `AXSTACK-DONE` plus terminal `t3_thread_wait` on the recorded run and candidate checks: non-empty diff, clean tree, and named red/green logs.
 Completion must match the current attempt key and candidate SHA; an older attempt never completes a newer one.
 Process the whole delivery before advancement, validating sender, scope,
 authority, revisions and artifacts. Deduplicate by native task/thread/run identity.
 A question marker stays incomplete; failure, interruption, permission prompts
 and refusals hold with preserved evidence. Receipt messages alone are progress.
+
+For a delegated child, detect questions with `t3_pending_request_list(childThreadId)`.
+Never rely on `hasPendingQuestion` alone.
+Codex `request_user_input_async` does not set it.
+Use `t3_pending_request_read`, then confirm the request is still pending with
+`t3_pending_request_list(childThreadId)`, then answer once within existing authority
+using `t3_pending_request_respond`; a second ask holds.
+Independent review rounds require fresh `delegate_task` calls.
+Only author continuation and an answer to a child's own question use the same child.
+Providers without a question tool (grok) and launched writers keep
+the `AXSTACK-QUESTION` marker route.
+
 Healthy unchanged passes are silent.
 An explicitly invoked phase dispatches its configured roles through T3 and
 closes with the lifecycle close-out; in-chat execution covers only ordinary
@@ -152,7 +167,7 @@ nothing without tested independent review.
 
 Close-out requires the [close-out acceptance table](run-record.md#close-out-acceptance).
 
-PRs merge by forge state; close out: (1) settle every T3 worker run and archive eligible threads. Settle each worker thread with `t3_thread_organize` and read back `settled: true`, because a finished task does not count; (2) compact record with counts and denominators—user
+PRs merge by forge state; close out: (1) settle every T3 worker run and archive eligible threads. Settle each worker thread with `t3_thread_organize` only if needed and read back `settled: true`, because a finished task does not count. For merged or closed PR authors, follow the [native/manual read-first path](t3-runtime.md#native-pr-links-and-watches); (2) compact record with counts and denominators—user
 interventions/deviations from plan/repairs; (3) `axstack-auditor`: settle
 non-zero/requested, else `counts zero`. A base auditor preflight rejection
 (no delegated task started) records `auditor: UNKNOWN (unlaunchable)` with the

package/skills/axstack/references/pr-shape.md

@@ -43,6 +43,21 @@ Smaller cohesive PRs are encouraged; no padding.
 
 ## PR body
 
+Open every own PR description with the problem in plain words from the user's
+request, then the fix.
+Do not lead with an implementation inventory.
+Use a title that explains why the change matters.
+For repository title conventions, check recently merged PRs.
+
+End the description with `Created with <model> in <harness>`.
+Include the author and subagent models that wrote code.
+Use the model name the harness reports.
+If the model is unknown, name the harness without guessing a model.
+
+Open real PRs, never drafts, so review bots run.
+If `gh pr create` fails because the GraphQL rate limit is exhausted, use REST:
+`gh api repos/<owner>/<repo>/pulls -f title=... -f head=... -f base=... -F body=@<file>`.
+
 Every own PR description must contain exactly one `Revert` line.
 Follow [Revert line](candidate-publication.md#revert-line) for its format and classification.
 

package/skills/axstack/references/pr-triage-nightly.md

@@ -1,8 +1,8 @@
 # Nightly PR-triage prompt
 
 You are a fresh read-only nightly PR-triage pass in your own T3 thread.
-Read-only applies to the forge, and own schedule replacement follows [Provider bindings](t3-runtime.md#preflight-and-binding).
-At pass start, follow [Provider bindings](t3-runtime.md#preflight-and-binding) for your own account and schedule binding.
+Read-only applies to the forge.
+At pass start, follow [Provider bindings](t3-runtime.md#preflight-and-binding) for provider binding.
 Use [Nightly PR-triage setup](automations.md#nightly-pr-triage-setup) for activation.
 Read the durable activation record for the repository set.
 Resolve the user with `gh api user --jq .login`.
@@ -20,3 +20,15 @@ Never declare a PR merge-ready.
 
 Never merge, post comments, change labels, push, send user messages, dispatch
 work, or launch threads.
+
+At pass start, read the recorded prior pass thread with `t3_thread_read` and
+apply [Finite-session teardown](automations.md#finite-session-teardown)'s guarded
+verification or hold/report before replacing its identity.
+Record own pass threadId in the existing activation record and read back the save.
+
+At pass end, an unbound pass calls `t3_thread_organize` with `action: settle` on its
+own thread only, only when it owns no PR watch, has zero unsettled descendants,
+has no pending receipt and is not held, wedged or a stalled owner.
+Follow [Finite-session teardown](automations.md#finite-session-teardown) for
+deferred self-settle and next-pass verification; this applies to quiet, skipped
+and empty passes too.

package/skills/axstack/references/review-manager-prompt.md

@@ -1,11 +1,5 @@
 # Review manager prompt
 
-The account picker must run first, before the quiet check: from the installed
-`axstack` skill directory, run `bun scripts/pick-instance.js --provider <own provider> --json`.
-Find this pass's current instance from `t3_thread_configuration` in the picker
-JSON `instances` and inspect its `state`; defer reference loads until a full pass.
-An eligible current instance remains quiet-eligible even when `chosenInstanceId`
-names a different instance.
 Run the quiet check before invoking `axstack-review` or reading any `skills/**` file.
 
 Resolve self with `gh api user --jq .login`, then union two paginated `gh search prs`
@@ -19,26 +13,32 @@ PR-job descendants, using exact lane identities.
 Read continuity at `~/.local/share/axstack/runs/review-manager/progress.md` and
 compare its Watermarks with the inventory:
 
-- Run the full pass when the picker JSON marks this pass's current instance `excluded`.
-- Run the full pass when the picker exits with code 1.
-- Run the full pass when the picker exits with code 2.
 - Run the full pass when the sorted list differs from `Quiet gate:`.
 - Run the full pass when any other lane thread is non-terminal.
+- Run the full pass only when the recorded prior pass thread is unsettled and no guard keeps it unsettled.
 - Run the full pass when continuity or either watermark is missing, malformed or unreadable.
 - Run the full pass when `Last full pass:` is at least 4 hours old or in the future.
 - Run the full pass when any call fails or pagination, PR head OIDs or thread/run liveness is incomplete or unknown.
 - Run the full pass when continuity records any unserved, capacity-deferred, `INCOMPLETE` or re-eligible event, or any open hold with a resume condition beyond a user decision (such as capacity or remeasure).
 
+Held or flagged prior passes never force a full pass or failure report by their unsettled state alone.
+
 Holds awaiting only the user remain quiet-eligible: a user decision on the PR
 changes `updatedAt`, and the 4 h full pass reconciles replies in flagged threads.
 
-If every busy condition is false, end the turn with no text, no skill load, no sweep and no continuity write.
+Before evaluating the quiet gate, read the recorded prior pass thread with
+`t3_thread_read`; unknown identity or readback forces a full pass.
+Record own pass threadId in review-manager continuity after reading the prior identity,
+before duplicate classification, then read back the identity save.
+
+If every busy condition is false, apply the guarded pass-end self-settle below,
+then end the turn with no text, no skill load, no sweep and no other continuity write.
 Otherwise continue with the full pass below; incomplete evidence retains the
 existing admission and cleanup holds.
 
 ## Full pass
 
-Follow [Provider bindings](t3-runtime.md#preflight-and-binding) for own account selection.
+Follow [Provider bindings](t3-runtime.md#preflight-and-binding) for provider binding.
 
 You are a fresh finite T3 review-manager pass in project `axstack-review-lane`
 on the VPS's existing `axatbhardwaj/axstack` clone. Fetch first; this unbound
@@ -60,3 +60,10 @@ The full pass records in Watermarks `Quiet gate:` as the sorted list and
 `Last full pass:` as UTC only as owner-of-record, then reads back the save.
 Settle descendants and end the finite turn. Stay quiet when unchanged.
 Never check out a PR branch here.
+
+At pass end, an unbound pass calls `t3_thread_organize` with `action: settle` on its
+own thread only, only when it owns no PR watch, has zero unsettled descendants,
+has no pending receipt and is not held, wedged or a stalled owner.
+Follow [Finite-session teardown](automations.md#finite-session-teardown) for
+deferred self-settle and next-pass verification; this applies to quiet, skipped
+and empty passes too.

package/skills/axstack/references/role-roster.md

@@ -18,8 +18,7 @@
   | `codex-only` | `codex/sol` | `axstack-reviewer-secondary` (`codex/luna` xhigh) |
   | `claude-only` | `claude/opus` | `axstack-reviewer-secondary` (`claude/sonnet` high) |
 - `axstack-advisor-astra`/`axstack-advisor-opus` advise and author candidates;
-  `axstack-arena-candidate-grok`/
-  `axstack-arena-candidate-antigravity` add families.
+  `axstack-arena-candidate-grok` adds a family.
   `axstack-arena-judge-opus` judges round 1; `axstack-escalation-fable`/`axstack-arena-judge-astra` judge round 2.
   High-stakes/trigger: fresh [contract](contracts.md) session.
   `axstack-auditor` audits; `axstack-checker` reports discrepancies.
@@ -39,11 +38,9 @@
 - `axstack-auditor`: `claude/haiku` xhigh in mixed, pinned to
   `claude-haiku-5-5`. Missing that exact model or xhigh effort holds without
   substitution. `claude/sonnet` high in claude-only; `codex/luna` xhigh in codex-only.
-- `axstack-explore-codebase`/`axstack-monitor`: `claude/haiku` high in
+- `axstack-explore-codebase`: `claude/haiku` high in
   mixed/claude-only, pinned to `claude-haiku-5-5`. Missing that exact model
-  or high effort holds these roles without substitution.
-  `axstack-monitor`: standalone watch never sends; chat-run watch: bounded
-  internal reports to its Run and original driver.
+  or high effort holds this role without substitution.
 - Scout requires `axstack-explore-codebase` and
   `axstack-explore-codebase-luna` per facet in mixed (Haiku high + Luna
   xhigh) and codex-only (Sol high + Luna xhigh). Claude-only requires
@@ -55,10 +52,8 @@
   cross-reading. The driver reconciles findings per claim, never averages.
   Record intentional absence and continue with Sonnet alone.
 - Optional seats: `axstack-arena-candidate-grok`,
-  `axstack-arena-candidate-antigravity`, `axstack-research-web-google`,
   `axstack-research-x`, `axstack-checker`, `axstack-research-code-sol`,
-  `axstack-explore-execution-sol`, `axstack-auditor-sol`,
-  `axstack-explore-codebase-flash`.
+  `axstack-explore-execution-sol`, `axstack-auditor-sol`.
   Always dispatch optional seats when configured and available, except research
   branches, which follow [Research step 2](../../axstack-research/SKILL.md#procedure) sizing.
   A research branch unneeded by an open claim is recorded `not needed (<reason>)`,

package/skills/axstack/references/routing.md

@@ -12,8 +12,7 @@ or use an explicit user selection in the run record. The legacy manifest
 manifest value does not create a contradiction. Missing or contradictory installed
 preset / explicit user selection: setup gap, hold. Never infer from live profiles, harness,
 tools, credentials, quota, subscription, or default to `mixed`.
-Only same-provider, same-model account selection among instances of one driver
-may use headroom under [Provider bindings](t3-runtime.md#preflight-and-binding);
+Follow [Provider bindings](t3-runtime.md#preflight-and-binding);
 provider/model substitution still requires the user's decision.
 
 At start, snapshot all role IDs from installed `skills/axstack/roles.json`,
@@ -30,9 +29,6 @@ using saved T3 capabilities JSON; missing or malformed capabilities holds.
 Claude exact IDs come from capabilities, replacing transcript read-back.
 Use an explicit model as given; for `model:null` without a class, grok uses the
 provider's first listed model from saved capabilities and records its exact ID.
-For `model:null` without a class, Antigravity must select the first listed model
-whose ID ends with `-<effort>` from saved capabilities and record its exact ID.
-Missing effort-suffix matches hold resolution for Antigravity.
 A Codex or Claude role with neither model nor class is an intentional absence
 and holds.
 Resume must reuse the saved capabilities and role snapshot with no re-resolution; changes require the user’s explicit decision.
@@ -40,8 +36,7 @@ An unavailable provider, model, role, mode or effort holds that role with no sub
 Preset changes apply to new runs only; an active run keeps its snapshot.
 Replacing a session needs an explicit user decision and revalidation.
 Timeout, quota, auth and rejection hold affected work.
-An exhausted account with no eligible sibling still holds; dispatched roles never fail over mid-thread.
-Follow [Provider bindings](t3-runtime.md#preflight-and-binding) for driver account re-selection at turn boundaries.
+Follow [Provider bindings](t3-runtime.md#preflight-and-binding).
 [Model discipline](contracts.md#model-discipline) governs optional seats,
 auditor preflight, and required holds; [Role roster](role-roster.md) governs
 single-provider absence, mixed research with at least two different providers,
@@ -148,4 +143,5 @@ not alone a formal spec trigger. Hold affected unsafe work while reassessing.
   author provenance — then use `axstack-review` and `axstack-watch` without
   repeated approval or new spec ceremony. Never infer the author from the
   orchestrator or assume an imported own PR's author.
+  Before pushing a stale or external PR, follow [adoption dispositions](../../axstack-watch/SKILL.md#stale-or-external-prs).
 - Direct later phase: start there and pass that phase's identity check.