@open-agent-toolkit/cli 0.2.24 → 0.2.26

package/assets/skills/oat-project-review-provide-remote/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: oat-project-review-provide-remote
-version: 1.0.4
+version: 1.1.0
 description: Use when reviewing a GitHub PR opened on another machine for an active OAT project and posting findings back as a single PR review. Resolves the project from the PR diff, reads project artifacts for mode-aware review, and posts via gh api.
 disable-model-invocation: true
 user-invocable: true
@@ -56,7 +56,7 @@ If you catch yourself:
 - Pushing or committing anything from this machine -> STOP; machine A's receive-remote owns all project-state mutations.
 - Posting the review to GitHub without explicit user confirmation -> STOP and present the body + verdict for approval first.
 - Narrowing against a prior review SHA without running the existence + ancestry guard -> STOP and run the guard first.
-- Narrowing against a prior review for a different project or a different scope -> STOP; only the same `(project, scope)` prior review narrows this one.
+- Narrowing against a prior review for a different project, scope, or lineage -> STOP; only the same `(project, scope, lineage)` prior review narrows this one.
 - Forgetting to remove the ephemeral worktree after posting (or on failure) -> STOP and release it in a `finally`.
 
 ## Progress Indicators (User-Facing)
@@ -90,7 +90,7 @@ oat-project-review-provide-remote [code <scope>|artifact <scope>]
 - `--pr <N>`: target PR number. When omitted, auto-detect from the current branch.
 - `--project <path>`: explicit OAT project directory. Takes precedence over the diff scan. Required when the diff touches zero or multiple projects' `state.md`.
 - `--no-checkout`: skip the ephemeral worktree and review from `gh pr diff` only (degraded context; project artifacts read from `gh` blob fetches instead of the checkout).
-- `--narrow` / `--no-narrow`: force or forbid re-review narrowing against a prior provide-remote review for this `(project, scope)`. When neither is passed, honor `workflow.autoNarrowReReviewScope` (no prompt when `true`; confirm prompt otherwise).
+- `--narrow` / `--no-narrow`: force or forbid re-review narrowing against a prior provide-remote review for this `(project, scope, lineage)`. `--narrow` takes precedence over the configured preference; `--no-narrow` forces full scope. When neither is passed, honor `workflow.autoNarrowReReviewScope`: unset and `true` narrow automatically, while only `false` forces full scope. This path never prompts for a narrowing decision.
 
 Inputs are CLI-style args parsed from `$ARGUMENTS`. No file inputs. No file outputs on this machine.
 
@@ -186,26 +186,69 @@ NEVER mutate, commit, or push any project artifact. This is a read-only project
 
 ### Step 4: Detect Prior Reviews + Narrow Scope
 
-List prior PR reviews and parse each body's marker block. Filter to reviews where `oat_provide_remote: true` AND `oat_project == "$PROJECT_PATH"` AND `oat_review_scope == "<current-scope-token>"`. Different-project or different-scope prior reviews do NOT narrow this one (mirrors the tested helper at `packages/cli/src/review-remote/narrowing.ts`, `pickNarrowingTarget` with `rail: "project"`).
+Resolve the current invocation lineage before selecting a candidate:
+
+- `manual` and `auto` are the lifecycle lineage and are interchangeable.
+- `gate` is a separate lineage qualified by the exact `oat_gate_target`.
+
+Resolve preference and per-invocation flags before candidate enumeration:
+
+1. `--no-narrow` forces full PR scope with reason `narrowing-disabled` and skips `gh api` review enumeration entirely.
+2. `--narrow` forces a narrowing attempt even when the configured preference is `false`.
+3. With neither flag, unset and `true` attempt narrowing; only `false` forces full PR scope with reason `narrowing-disabled` and skips `gh api` review enumeration entirely.
+
+Only when narrowing will be attempted, list prior PR reviews and parse each body's target-qualified GitHub review marker block. Capture command status and stderr separately:
 
 ```bash
-gh api "/repos/{owner}/{repo}/pulls/$PR/reviews"
+REVIEWS_ERROR_FILE=""
+REVIEWS_DIAGNOSTIC=""
+REVIEWS_DISCOVERY_OK=false
+
+if ! REVIEWS_ERROR_FILE=$(mktemp "${TMPDIR:-/tmp}/oat-review-errors.XXXXXX"); then
+  REVIEWS_DIAGNOSTIC="diagnostic-file-unavailable: unable to create a temporary stderr file"
+elif REVIEWS_JSON=$(gh api "/repos/{owner}/{repo}/pulls/$PR/reviews" 2>"$REVIEWS_ERROR_FILE"); then
+  REVIEWS_DISCOVERY_OK=true
+  rm -f -- "$REVIEWS_ERROR_FILE"
+  REVIEWS_ERROR_FILE=""
+else
+  REVIEWS_DIAGNOSTIC=$(dd if="$REVIEWS_ERROR_FILE" bs=500 count=1 2>/dev/null)
+  rm -f -- "$REVIEWS_ERROR_FILE"
+  REVIEWS_ERROR_FILE=""
+fi
 ```
 
-Take the most recent matching review (by submitted timestamp). Before narrowing to `<prior_sha>..<HEAD>`, run the stale-SHA guard (design.md → Error Handling → Stale prior-review SHA):
+Do not run `gh api` when diagnostic-file creation fails. Treat that failure, a nonzero `gh api` result, or a response-level enumeration/parsing failure as discovery failure. Before removing the diagnostic file, preserve at most 500 bytes from stderr. If response parsing later fails, set `REVIEWS_DISCOVERY_OK=false` and put at most 500 characters of the parse error in `REVIEWS_DIAGNOSTIC`. Apply the forced/automatic policy below whenever `REVIEWS_DISCOVERY_OK=false`. Do not treat individual irrelevant or lineage-ineligible marker blocks as response-level failure.
+
+- With forced `--narrow`, discovery failure is a hard error with reason `prior-reviews-unavailable`; report the diagnostic and stop instead of pretending to narrow.
+- On the automatic path, discovery failure fails open to full PR scope with stable reason `prior-reviews-unavailable` plus the diagnostic detail, then continues to review the full PR diff.
+
+After successful enumeration, candidate discovery belongs to this rail: use only marker blocks where `oat_provide_remote: true`, `oat_project == "$PROJECT_PATH"`, and `oat_review_scope == "<current-scope-token>"`, then require the same lineage. A lifecycle invocation accepts only a prior `manual` or `auto` marker. A gate invocation accepts only a prior `gate` marker carrying the same exact non-empty `oat_gate_target`; a target-less gate marker is ineligible, and a gate never inherits a lifecycle review or another gate target. Invocation-less or unknown legacy markers, and markers for another project or scope, remain unknown and ineligible. Do not read the local lifecycle Reviews table or use a project-plan fallback on this rail.
+
+Take the most recent eligible marker by submitted timestamp. Candidate discovery and marker provenance remain rail-specific; only after this rail supplies that candidate apply the shared guard, fallback, and classification semantics mirrored by `packages/cli/src/review-remote/narrowing.ts`. When no eligible marker supplies a full 40-character hexadecimal `oat_review_head_sha`, use full PR scope with reason `no-prior-review`. Do not run Git guards for an invalid candidate.
+
+Before accepting `<prior_sha>..<PR_HEAD_SHA>`, run the stale-SHA guard (design.md → Error Handling → Stale prior-review SHA):
 
 Run the guard in the available git context `$GIT_CTX` — `$EPHEMERAL_PATH` in rich-context (checkout) mode, or `$REPO_ROOT` in diff-only mode (where no worktree exists):
 
 1. **Existence:** `git -C "$GIT_CTX" cat-file -e <prior_sha>` (diff-only mode: `git -C "$REPO_ROOT" fetch origin <prior_sha>:refs/oat-prior-review` first, then re-check; if that fetch fails, fall back to full PR scope).
 2. **Ancestry:** `git -C "$GIT_CTX" merge-base --is-ancestor <prior_sha> "$PR_HEAD_SHA"`.
 
-Guard outcomes:
+Guard outcomes and reasons:
 
-- Both pass -> narrow to `<prior_sha>..<HEAD>`.
-- Either fails -> fall back to full PR scope and warn that the prior SHA is unreachable (likely rebase/force-push).
+- Both pass -> narrow to exactly `<prior_sha>..<PR_HEAD_SHA>` with reason `narrowed from guarded prior reviewed head`.
+- Object missing or fetch failure -> fall back to full PR scope with reason `stale-sha` and detail `prior-commit-missing`.
+- Ancestry failure -> fall back to full PR scope with reason `stale-sha` and detail `prior-commit-not-ancestor`.
 - `--narrow` set AND guard fails -> hard error; surface unreachability and stop.
-- `workflow.autoNarrowReReviewScope == true` -> never prompt; guard failure auto-falls back to full scope with the warning as the auto-fallback notice.
-- No matching prior review (for this `(project, scope)`) -> use full PR diff.
+- The automatic path fails open for every candidate or guard failure and preserves the applicable reason.
+
+When a guarded narrow range is accepted, enumerate its changed files with `git diff --name-only "<prior_sha>..<PR_HEAD_SHA>"` and classify it for reporting. Resolve `PROJECT_RELATIVE` as the active project's repository-relative path first:
+
+- `empty`: no changed files.
+- `bookkeeping-only`: every changed path is inside the exact active project prefix `"$PROJECT_RELATIVE/"`.
+- `substantive`: any changed path is outside that exact prefix.
+- If changed-file enumeration fails, conservatively classify as `substantive` and add `changed-files-unavailable` to the reason.
+
+Classification is reporting-only: `empty`, `bookkeeping-only`, and `substantive` all dispatch the review over the same resolved range. It never skips, gates, or shortens the review. Print one narrowing resolution line containing the resolved full-scope or narrow range, classification when narrowed, and the explicit reason. Do not print another narrowing prompt or decision line elsewhere.
 
 ### Step 5: Run the Review (Tier 1/2/3 Dispatch)
 
@@ -311,11 +354,12 @@ oat_provide_remote: true
 oat_review_head_sha: <PR_HEAD_SHA>
 oat_review_scope: <scope token, e.g. p02 | final>
 oat_project: <PROJECT_PATH>
-oat_review_invocation: manual
+oat_review_invocation: <manual | auto | gate>
+oat_gate_target: <exact configured gate target; omit for lifecycle>
 -->
 ```
 
-`oat_project` carries the resolved project path so machine A's `oat-project-review-receive-remote` routes findings into the right project's plan tasks. Verdict: `REQUEST_CHANGES` when any critical or important finding exists; `COMMENT` otherwise (including a clean, zero-findings review — never auto-`APPROVE`).
+`oat_project` carries the resolved project path so machine A's `oat-project-review-receive-remote` routes findings into the right project's plan tasks. Preserve the current invocation lineage in the GitHub marker: omit `oat_gate_target` for lifecycle reviews and include the exact target for gate reviews. Verdict: `REQUEST_CHANGES` when any critical or important finding exists; `COMMENT` otherwise (including a clean, zero-findings review — never auto-`APPROVE`).
 
 ### Step 8: Post the Review + Clean Up
 
@@ -353,6 +397,9 @@ Posting failure handling (design.md → Error Handling → Posting failures): on
 Always release the ephemeral worktree in a `finally`, even when review or posting fails:
 
 ```bash
+if [[ -n "${REVIEWS_ERROR_FILE:-}" ]]; then
+  rm -f -- "$REVIEWS_ERROR_FILE"
+fi
 git -C "$REPO_ROOT" worktree remove --force "$EPHEMERAL_PATH" || git -C "$REPO_ROOT" worktree prune
 rm -rf "$EPHEMERAL_PATH"
 ```
@@ -378,7 +425,7 @@ At completion, report:
 - Review type and scope token.
 - Read mode (worktree checkout vs diff-only).
 - Dispatch tier used (Tier 1 structured-output / Tier 2 fresh session / Tier 3 inline).
-- Narrowing decision (full scope vs `<prior_sha>..<HEAD>`, and why).
+- Narrowing decision (full scope vs `<prior_sha>..<PR_HEAD_SHA>`), reporting-only classification when narrowed, and the explicit reason.
 - Severity counts and total findings.
 - Inline-comment count posted vs findings downgraded to the body (out-of-diff).
 - Verdict (`REQUEST_CHANGES` or `COMMENT`).
@@ -391,7 +438,9 @@ At completion, report:
 - PR scope resolved and confirmed.
 - OAT project resolved from the diff (or `--project` override) and validated; project artifacts read read-only for mode-aware context.
 - PR content acquired via ephemeral worktree (or diff-only fallback) without mutating the caller's working tree.
-- Prior provide-remote reviews detected and filtered to this `(project, scope)`; re-review narrowing applied only after the stale-SHA guard passes.
+- Prior provide-remote GitHub markers detected and filtered to this `(project, scope, lineage)`; re-review narrowing applied only after the full-SHA, existence, and ancestry guards pass, with explicit fail-open reasons and no local Reviews-table or project-plan fallback.
+- Unset and `true` narrow automatically, only `false` forces full scope, per-invocation flags retain precedence, and no narrowing prompt remains.
+- Narrowed ranges are classified as `empty`, `bookkeeping-only`, or `substantive` for reporting only; changed-file enumeration failure conservatively reports `substantive`.
 - Review executed via Tier 1 (`oat-reviewer` structured-output, NO artifact) /
   Tier 2 (fresh session) / Tier 3 (inline), with fallthrough limited to
   dispatch unavailability or explicit pre-start native role rejection.

package/assets/skills/oat-project-review-receive/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: oat-project-review-receive
-version: 1.5.9
+version: 1.6.0
 description: Use when the user explicitly asks to receive review findings for an OAT project — e.g. "receive review", "process review", "process the project review", or confirms a previously offered review-receive step. Do NOT auto-invoke merely because a review file exists. Resolves the latest review and offers before acting.
 disable-model-invocation: false
 user-invocable: true
@@ -139,6 +139,14 @@ Use this fallback only for active project reviews. It cannot discover ad-hoc rev
 
 **Read the selected review file completely.**
 
+For a code review, read `oat_review_head_sha`, `oat_review_invocation`, and
+`oat_gate_target` from frontmatter. Accept the reviewed head only when it is a
+full 40-character hexadecimal SHA; never expand or infer an abbreviated,
+symbolic, or range value during receive. A missing invocation remains unknown;
+do not assume a legacy artifact was manual. Write `-` for absent provenance
+rather than borrowing values from another review event. `oat_gate_target` is
+meaningful only when the invocation is `gate`.
+
 Derive archive bookkeeping before making lifecycle edits:
 
 ```bash
@@ -406,8 +414,17 @@ Add new tasks to plan.md in the target phase. When adding or editing tasks, pres
 **Review-fix bookkeeping (required):**
 - When you add review-generated fix tasks:
   - Locate the Reviews event matching the selected review's Scope, Type, and `SOURCE_REVIEW_FILENAME`, then update it to `fixes_added` (work queued), set the Date, and replace its Artifact with `reviews/archived/$REVIEW_FILENAME`.
+  - For code events, populate or preserve `Reviewed Head`, `Invocation`, and
+    `Gate Target` from the selected artifact's validated frontmatter. Never
+    replace known provenance with `-`; if the artifact is legacy or a value is
+    invalid, preserve existing cells and otherwise leave the value unknown.
   - The written `REVIEW_FILENAME` becomes the event's artifact filename and identity for every later mutation; use the already-resolved final basename in every plan and implementation reference.
   - Never select a row by scope alone or move an event status backward. If the exact bound event is missing, stop and reconcile the ledger instead of mutating another event.
+  - Mutate cells by header name. If the table is still the legacy five-column
+    shape, add `Reviewed Head`, `Invocation`, and `Gate Target` to the header
+    and separator and pad every existing row with `-`. If it is already
+    widened, preserve all existing and unknown trailing cells; never rebuild a
+    row as only five or eight cells.
   - Update `## Implementation Complete` totals (phase counts + total task count) so downstream PR/review summaries don’t go stale.
   - If the plan includes any phase rollups that reference task counts, update those too.
 
@@ -429,6 +446,9 @@ Add new tasks to plan.md in the target phase. When adding or editing tasks, pres
   - Status: `fixes_added` (if tasks were added) or `passed` (if no Critical/Important/Medium and no unresolved final-scope gates)
   - Date: `{today}`
   - Artifact: `reviews/archived/$REVIEW_FILENAME`
+  - Reviewed Head: validated full `oat_review_head_sha` for code reviews
+  - Invocation: `oat_review_invocation` for code reviews
+  - Gate Target: exact `oat_gate_target` for gate code reviews; `-` otherwise
 ````
 
 **Status semantics (v1):**

package/assets/skills/oat-project-review-receive-remote/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: oat-project-review-receive-remote
-version: 1.4.2
+version: 1.5.0
 description: Use when processing GitHub PR review comments within project context. Fetches PR comments, creates plan tasks, and updates project artifacts.
 disable-model-invocation: true
 user-invocable: true
@@ -115,6 +115,33 @@ Confirm resolved PR number with user.
 npx agent-reviews --json --unresolved --pr <N>
 ```
 
+Resolve provenance from the GitHub review event that produced the fetched
+feedback, not from the PR's current head. Parse its OAT marker block when
+present:
+
+- accept `oat_review_head_sha` only as a full 40-character hexadecimal SHA;
+- preserve `oat_review_invocation` as the invocation kind;
+- preserve `oat_gate_target` only for a gate invocation.
+
+If the source review is legacy, multiple source reviews disagree, or the
+lineage cannot be associated with this receive event, leave the corresponding
+ledger cells unknown (`-`). Never infer lineage from the reviewer identity or
+substitute the current PR head for the commit that was actually reviewed.
+
+**Reviews ledger write contract (all receive paths):**
+
+- Resolve `Scope`, `Type`, `Status`, `Date`, `Artifact`, `Reviewed Head`,
+  `Invocation`, and `Gate Target` by header name before mutating a row; never
+  use fixed cell positions.
+- If the Reviews table has only the legacy five columns, add `Reviewed Head`,
+  `Invocation`, and `Gate Target` to its header and separator and pad every
+  existing row with `-`. In an already widened table, pad a shorter row with
+  `-` through the current header width before mutation.
+- Mutate only the event selected by the event-identity rules below. Preserve
+  every unknown column in its original position and every existing known value
+  unless the operation explicitly advances that cell. Never truncate a row to
+  five, eight, or any other assumed width.
+
 If no unresolved comments:
 
 1. Create a UTC timestamp and event-distinct filename:
@@ -125,10 +152,14 @@ If no unresolved comments:
    remain in top-level `reviews/`.
 3. Record the clean result as a `passed` Reviews event whose event identity
    combines `Scope`, `Type`, and artifact filename:
+   - Apply the Reviews ledger write contract above before claiming, appending,
+     or mutating the event.
    - Claim only an unbound `pending` placeholder with matching Scope + Type and
      Artifact `-`; otherwise append a distinct row.
    - Set Date and Artifact to this clean event. Advance only this event and
      never mutate another row by scope alone.
+   - Populate `Reviewed Head`, `Invocation`, and `Gate Target` only from the
+     validated source-review provenance above; use `-` when it is unknown.
 4. Commit `plan.md` and the clean review artifact atomically with
    `chore(oat): record clean remote review (pr-#<N>)`. Do not stop with
    uncommitted bookkeeping.
@@ -176,7 +207,8 @@ For each converted finding:
 ### Step 6: Update Project Artifacts
 
 Before changing the ledger, write an event-distinct review artifact containing
-the PR number, fetch timestamp, normalized findings, and dispositions:
+the PR number, fetch timestamp, normalized findings, dispositions, and any
+validated source-review head/invocation/gate-target provenance:
 `reviews/archived/remote-pr-<N>-review-YYYY-MM-DDTHHMMSSZ.md`. Remote receive
 fully dispositions the event as `passed` or `fixes_added`, so the artifact is
 consumed immediately and belongs in `reviews/archived/`, not top-level
@@ -198,11 +230,16 @@ Update `plan.md`:
   - status `passed` when no actionable findings remain
   - date set to today
   - artifact `reviews/archived/remote-pr-<N>-review-YYYY-MM-DDTHHMMSSZ.md`
+  - reviewed head set to the validated full source-review SHA, or `-`
+  - invocation set to the source-review invocation kind, or `-`
+  - gate target set only for a gate source review, or `-`
 - Claim an unbound `pending` placeholder only when its Scope + Type matches and
   its Artifact is `-`; otherwise append the event. Later mutations select it by
   Scope + Type + artifact filename, never by scope or `github-pr #<N>` alone.
 - Never move an event status backward or overwrite an earlier event from the
   same PR.
+- Apply the Reviews ledger write contract above before claiming, appending, or
+  mutating the event.
 - Update `## Implementation Complete` totals.
 
 Update `implementation.md`:

package/assets/skills/oat-review-provide-remote/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: oat-review-provide-remote
-version: 1.0.3
+version: 1.1.0
 description: Use when reviewing a GitHub PR opened on another machine and posting findings back as a single PR review, outside any OAT project context. Fetches the PR via gh, reviews it, and posts via gh api.
 disable-model-invocation: true
 user-invocable: true
@@ -77,7 +77,7 @@ oat-review-provide-remote [--pr <N>] [--no-checkout] [--narrow|--no-narrow]
 
 - `--pr <N>`: target PR number. When omitted, auto-detect from the current branch.
 - `--no-checkout`: skip the ephemeral worktree and review from `gh pr diff` only (degraded context).
-- `--narrow` / `--no-narrow`: force or forbid re-review narrowing against a prior provide-remote review. When neither is passed, honor `workflow.autoNarrowReReviewScope` (no prompt when `true`; confirm prompt otherwise).
+- `--narrow` / `--no-narrow`: force or forbid re-review narrowing against a prior ad-hoc provide-remote review. `--narrow` takes precedence over the configured preference; `--no-narrow` forces full scope. When neither is passed, honor `workflow.autoNarrowReReviewScope`: unset and `true` narrow automatically, while only `false` forces full scope. This path never prompts for a narrowing decision.
 
 Inputs are CLI-style args parsed from `$ARGUMENTS`. No file inputs. No file outputs on this machine.
 
@@ -143,26 +143,63 @@ gh pr view "$PR" --json title,body,headRefOid,baseRefName,state
 
 ### Step 3: Detect Prior Reviews + Narrow Scope
 
-List prior PR reviews and parse each body's marker block. Filter to reviews where `oat_provide_remote: true` AND `oat_review_scope == "ad-hoc"` AND no `oat_project` key — project-rail markers are ignored by the ad-hoc filter.
+Resolve preference and per-invocation flags before candidate enumeration:
+
+1. `--no-narrow` forces full PR scope with reason `narrowing-disabled` and skips `gh api` review enumeration entirely.
+2. `--narrow` forces a narrowing attempt even when the configured preference is `false`.
+3. With neither flag, unset and `true` attempt narrowing; only `false` forces full PR scope with reason `narrowing-disabled` and skips `gh api` review enumeration entirely.
+
+Only when narrowing will be attempted, list prior PR reviews and parse each body's GitHub review marker block. Capture command status and stderr separately:
 
 ```bash
-gh api "/repos/{owner}/{repo}/pulls/$PR/reviews"
+REVIEWS_ERROR_FILE=""
+REVIEWS_DIAGNOSTIC=""
+REVIEWS_DISCOVERY_OK=false
+
+if ! REVIEWS_ERROR_FILE=$(mktemp "${TMPDIR:-/tmp}/oat-review-errors.XXXXXX"); then
+  REVIEWS_DIAGNOSTIC="diagnostic-file-unavailable: unable to create a temporary stderr file"
+elif REVIEWS_JSON=$(gh api "/repos/{owner}/{repo}/pulls/$PR/reviews" 2>"$REVIEWS_ERROR_FILE"); then
+  REVIEWS_DISCOVERY_OK=true
+  rm -f -- "$REVIEWS_ERROR_FILE"
+  REVIEWS_ERROR_FILE=""
+else
+  REVIEWS_DIAGNOSTIC=$(dd if="$REVIEWS_ERROR_FILE" bs=500 count=1 2>/dev/null)
+  rm -f -- "$REVIEWS_ERROR_FILE"
+  REVIEWS_ERROR_FILE=""
+fi
 ```
 
-Take the most recent matching review (by submitted timestamp). Before narrowing to `<prior_sha>..<HEAD>`, run the stale-SHA guard (design.md → Error Handling → Stale prior-review SHA):
+Do not run `gh api` when diagnostic-file creation fails. Treat that failure, a nonzero `gh api` result, or a response-level enumeration/parsing failure as discovery failure. Before removing the diagnostic file, preserve at most 500 bytes from stderr. If response parsing later fails, set `REVIEWS_DISCOVERY_OK=false` and put at most 500 characters of the parse error in `REVIEWS_DIAGNOSTIC`. Apply the forced/automatic policy below whenever `REVIEWS_DISCOVERY_OK=false`. Do not treat individual irrelevant or lineage-ineligible marker blocks as response-level failure.
+
+- With forced `--narrow`, discovery failure is a hard error with reason `prior-reviews-unavailable`; report the diagnostic and stop instead of pretending to narrow.
+- On the automatic path, discovery failure fails open to full PR scope with stable reason `prior-reviews-unavailable` plus the diagnostic detail, then continues to review the full PR diff.
+
+After successful enumeration, candidate discovery belongs to this rail: use only marker blocks where `oat_provide_remote: true`, `oat_review_scope == "ad-hoc"`, no `oat_project` key is present, and the invocation is in the ad-hoc lifecycle lineage (`manual` or `auto`). Project-rail, gate, and invocation-less or unknown legacy markers remain unknown and ineligible. This rail has no project-plan or local lifecycle Reviews-table fallback.
+
+Take the most recent eligible marker by submitted timestamp. Candidate discovery and marker provenance remain rail-specific; only after this rail supplies that candidate apply the shared guard, fallback, and classification semantics mirrored by `packages/cli/src/review-remote/narrowing.ts`. When no eligible marker supplies a full 40-character hexadecimal `oat_review_head_sha`, use full PR scope with reason `no-prior-review`. Do not run Git guards for an invalid candidate.
+
+Before accepting `<prior_sha>..<PR_HEAD_SHA>`, run the stale-SHA guard (design.md → Error Handling → Stale prior-review SHA):
 
 Run the guard in the available git context `$GIT_CTX` — `$EPHEMERAL_PATH` in rich-context (checkout) mode, or `$REPO_ROOT` in diff-only mode (where no worktree exists):
 
 1. **Existence:** `git -C "$GIT_CTX" cat-file -e <prior_sha>` (diff-only mode: `git -C "$REPO_ROOT" fetch origin <prior_sha>:refs/oat-prior-review` first, then re-check; if that fetch fails, fall back to full PR scope).
 2. **Ancestry:** `git -C "$GIT_CTX" merge-base --is-ancestor <prior_sha> "$PR_HEAD_SHA"`.
 
-Guard outcomes:
+Guard outcomes and reasons:
 
-- Both pass -> narrow to `<prior_sha>..<HEAD>`.
-- Either fails -> fall back to full PR scope and warn that the prior SHA is unreachable (likely rebase/force-push).
+- Both pass -> narrow to exactly `<prior_sha>..<PR_HEAD_SHA>` with reason `narrowed from guarded prior reviewed head`.
+- Object missing or fetch failure -> fall back to full PR scope with reason `stale-sha` and detail `prior-commit-missing`.
+- Ancestry failure -> fall back to full PR scope with reason `stale-sha` and detail `prior-commit-not-ancestor`.
 - `--narrow` set AND guard fails -> hard error; surface unreachability and stop.
-- `workflow.autoNarrowReReviewScope == true` -> never prompt; guard failure auto-falls back to full scope with the warning as the auto-fallback notice.
-- No matching prior review -> use full PR diff.
+- The automatic path fails open for every candidate or guard failure and preserves the applicable reason.
+
+When a guarded narrow range is accepted, enumerate its changed files with `git diff --name-only "<prior_sha>..<PR_HEAD_SHA>"` and classify it for reporting:
+
+- `empty`: no changed files.
+- `substantive`: any non-empty range. The ad-hoc rail has no active project prefix, so it cannot classify a range as `bookkeeping-only`.
+- If changed-file enumeration fails, conservatively classify as `substantive` and add `changed-files-unavailable` to the reason.
+
+The shared vocabulary remains `empty`, `bookkeeping-only`, or `substantive`; `bookkeeping-only` is reachable only on a project rail with an exact active-project prefix. Classification is reporting-only: it never skips, gates, or shortens the review. Print one narrowing resolution line containing the resolved full-scope or narrow range, classification when narrowed, and the explicit reason. Do not print another narrowing prompt or decision line elsewhere.
 
 ### Step 4: Run the Review (Inline)
 
@@ -233,6 +270,9 @@ Posting failure handling (design.md → Error Handling → Posting failures): on
 Always release the ephemeral worktree in a `finally`, even when review or posting fails:
 
 ```bash
+if [[ -n "${REVIEWS_ERROR_FILE:-}" ]]; then
+  rm -f -- "$REVIEWS_ERROR_FILE"
+fi
 git -C "$REPO_ROOT" worktree remove --force "$EPHEMERAL_PATH" || git -C "$REPO_ROOT" worktree prune
 rm -rf "$EPHEMERAL_PATH"
 ```
@@ -252,7 +292,7 @@ At completion, report:
 
 - PR number and HEAD SHA reviewed.
 - Read mode (worktree checkout vs diff-only).
-- Narrowing decision (full scope vs `<prior_sha>..<HEAD>`, and why).
+- Narrowing decision (full scope vs `<prior_sha>..<PR_HEAD_SHA>`), reporting-only classification when narrowed, and the explicit reason.
 - Severity counts and total findings.
 - Inline-comment count posted vs findings downgraded to the body (out-of-diff).
 - Verdict (`REQUEST_CHANGES` or `COMMENT`).
@@ -263,7 +303,10 @@ At completion, report:
 
 - PR scope resolved and confirmed.
 - PR content acquired via ephemeral worktree (or diff-only fallback) without mutating the caller's working tree.
-- Prior provide-remote reviews detected; re-review narrowing applied only after the stale-SHA guard passes.
+- Prior ad-hoc GitHub markers detected and restricted to the ad-hoc lifecycle lineage; project and gate markers are ineligible, and no project-plan or local Reviews-table fallback is used.
+- Re-review narrowing applies only after the full-SHA, existence, and ancestry guards pass, with explicit fail-open reasons.
+- Unset and `true` narrow automatically, only `false` forces full scope, per-invocation flags retain precedence, and no narrowing prompt remains.
+- Narrowed ranges are classified as `empty` or `substantive` for reporting only; changed-file enumeration failure conservatively reports `substantive`.
 - Findings produced with consistent 4-tier severities and file:line references.
 - Inline comments mapped to in-diff positions; out-of-diff findings downgraded to the body, never dropped.
 - Posted-review body carries the marker block first, correct severity counts, and the minor-fix nudge when minors are present.

package/assets/templates/plan.md

@@ -168,13 +168,19 @@ git commit -m "feat(p01-t02): {description}"
 
 {Keep both code + artifact rows below. Add additional code rows (p03, p04, etc.) as needed, but do not delete `spec`/`design`.}
 
-| Scope  | Type     | Status  | Date | Artifact |
-| ------ | -------- | ------- | ---- | -------- |
-| p01    | code     | pending | -    | -        |
-| p02    | code     | pending | -    | -        |
-| final  | code     | pending | -    | -        |
-| spec   | artifact | pending | -    | -        |
-| design | artifact | pending | -    | -        |
+| Scope  | Type     | Status  | Date | Artifact | Reviewed Head | Invocation | Gate Target |
+| ------ | -------- | ------- | ---- | -------- | ------------- | ---------- | ----------- |
+| p01    | code     | pending | -    | -        | -             | -          | -           |
+| p02    | code     | pending | -    | -        | -             | -          | -           |
+| final  | code     | pending | -    | -        | -             | -          | -           |
+| spec   | artifact | pending | -    | -        | -             | -          | -           |
+| design | artifact | pending | -    | -        | -             | -          | -           |
+
+For code-review events, `Reviewed Head` is the full 40-character SHA at the
+head of the reviewed range. `Invocation` records `manual`, `auto`, or `gate`;
+`Gate Target` is populated only for gate events. Legacy five-column rows remain
+valid. Writers must preserve every existing row and every unknown trailing
+cell; never truncate a widened row back to five columns.
 
 **Status values:** `pending` → `received` → `fixes_added` → `fixes_completed` → `passed`
 

package/dist/commands/config/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/commands/config/index.ts"],"names":[],"mappings":"AAGA,OAAO,EAAE,mBAAmB,EAAE,KAAK,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAEhF,OAAO,EAEL,KAAK,aAAa,EACnB,MAAM,iCAAiC,CAAC;AAezC,OAAO,EAML,KAAK,SAAS,EACd,KAAK,cAAc,EAGnB,KAAK,UAAU,EAehB,MAAM,oBAAoB,CAAC;AAC5B,OAAO,EAEL,KAAK,cAAc,EAEpB,MAAM,iBAAiB,CAAC;AAGzB,OAAO,EAGL,KAAK,8BAA8B,EACnC,KAAK,yBAAyB,EAC/B,MAAM,kCAAkC,CAAC;AAC1C,OAAO,EACL,mCAAmC,EACnC,0BAA0B,EAC3B,MAAM,yCAAyC,CAAC;AACjD,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AAoGpC,UAAU,yBAAyB;IACjC,mBAAmB,EAAE,CACnB,OAAO,EAAE,UAAU,CAAC,OAAO,mBAAmB,CAAC,CAAC,CAAC,CAAC,KAC/C,cAAc,CAAC;IACpB,kBAAkB,EAAE,CAAC,GAAG,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IACrD,aAAa,EAAE,CAAC,QAAQ,EAAE,MAAM,KAAK,OAAO,CAAC,SAAS,CAAC,CAAC;IACxD,cAAc,EAAE,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,EAAE,SAAS,KAAK,OAAO,CAAC,IAAI,CAAC,CAAC;IACvE,kBAAkB,EAAE,CAAC,QAAQ,EAAE,MAAM,KAAK,OAAO,CAAC,cAAc,CAAC,CAAC;IAClE,mBAAmB,EAAE,CACnB,QAAQ,EAAE,MAAM,EAChB,MAAM,EAAE,cAAc,KACnB,OAAO,CAAC,IAAI,CAAC,CAAC;IACnB,cAAc,EAAE,CAAC,aAAa,EAAE,MAAM,KAAK,OAAO,CAAC,UAAU,CAAC,CAAC;IAC/D,eAAe,EAAE,CAAC,aAAa,EAAE,MAAM,EAAE,MAAM,EAAE,UAAU,KAAK,OAAO,CAAC,IAAI,CAAC,CAAC;IAC9E,mBAAmB,EAAE,CACnB,QAAQ,EAAE,MAAM,EAChB,GAAG,EAAE,MAAM,CAAC,UAAU,KACnB,OAAO,CAAC,MAAM,CAAC,CAAC;IACrB,sBAAsB,EAAE,CACtB,QAAQ,EAAE,MAAM,EAChB,aAAa,EAAE,MAAM,EACrB,GAAG,EAAE,MAAM,CAAC,UAAU,KACnB,OAAO,CAAC,cAAc,CAAC,CAAC;IAC7B,iBAAiB,EAAE,MAAM,OAAO,CAAC,MAAM,CAAC,CAAC;IACzC,QAAQ,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IAC5C,aAAa,EAAE,CAAC,OAAO,EAAE,MAAM,EAAE,GAAG,EAAE,aAAa,KAAK,OAAO,CAAC,OAAO,CAAC,CAAC;IACzE,kBAAkB,EAAE,CAClB,QAAQ,EAAE,MAAM,EAChB,KAAK,EAAE,MAAM,EACb,OAAO,EAAE,yBAAyB,KAC/B,OAAO,CAAC,8BAA8B,CAAC,CAAC;IAC7C,mCAAmC,EAAE,OAAO,mCAAmC,CAAC;IAChF,0BAA0B,EAAE,OAAO,0BAA0B,CAAC;IAC9D,UAAU,EAAE,MAAM,CAAC,UAAU,CAAC;CAC/B;AA63ED,wBAAgB,mBAAmB,CACjC,SAAS,GAAE,OAAO,CAAC,yBAAyB,CAAM,GACjD,OAAO,CAsKT"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/commands/config/index.ts"],"names":[],"mappings":"AAGA,OAAO,EAAE,mBAAmB,EAAE,KAAK,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAEhF,OAAO,EAEL,KAAK,aAAa,EACnB,MAAM,iCAAiC,CAAC;AAmBzC,OAAO,EAML,KAAK,SAAS,EACd,KAAK,cAAc,EAGnB,KAAK,UAAU,EAehB,MAAM,oBAAoB,CAAC;AAC5B,OAAO,EAEL,KAAK,cAAc,EAEpB,MAAM,iBAAiB,CAAC;AAGzB,OAAO,EAGL,KAAK,8BAA8B,EACnC,KAAK,yBAAyB,EAC/B,MAAM,kCAAkC,CAAC;AAE1C,OAAO,EACL,mCAAmC,EACnC,0BAA0B,EAC3B,MAAM,yCAAyC,CAAC;AACjD,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AAoGpC,UAAU,yBAAyB;IACjC,mBAAmB,EAAE,CACnB,OAAO,EAAE,UAAU,CAAC,OAAO,mBAAmB,CAAC,CAAC,CAAC,CAAC,KAC/C,cAAc,CAAC;IACpB,kBAAkB,EAAE,CAAC,GAAG,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IACrD,aAAa,EAAE,CAAC,QAAQ,EAAE,MAAM,KAAK,OAAO,CAAC,SAAS,CAAC,CAAC;IACxD,cAAc,EAAE,CAAC,QAAQ,EAAE,MAAM,EAAE,MAAM,EAAE,SAAS,KAAK,OAAO,CAAC,IAAI,CAAC,CAAC;IACvE,kBAAkB,EAAE,CAAC,QAAQ,EAAE,MAAM,KAAK,OAAO,CAAC,cAAc,CAAC,CAAC;IAClE,mBAAmB,EAAE,CACnB,QAAQ,EAAE,MAAM,EAChB,MAAM,EAAE,cAAc,KACnB,OAAO,CAAC,IAAI,CAAC,CAAC;IACnB,cAAc,EAAE,CAAC,aAAa,EAAE,MAAM,KAAK,OAAO,CAAC,UAAU,CAAC,CAAC;IAC/D,eAAe,EAAE,CAAC,aAAa,EAAE,MAAM,EAAE,MAAM,EAAE,UAAU,KAAK,OAAO,CAAC,IAAI,CAAC,CAAC;IAC9E,mBAAmB,EAAE,CACnB,QAAQ,EAAE,MAAM,EAChB,GAAG,EAAE,MAAM,CAAC,UAAU,KACnB,OAAO,CAAC,MAAM,CAAC,CAAC;IACrB,sBAAsB,EAAE,CACtB,QAAQ,EAAE,MAAM,EAChB,aAAa,EAAE,MAAM,EACrB,GAAG,EAAE,MAAM,CAAC,UAAU,KACnB,OAAO,CAAC,cAAc,CAAC,CAAC;IAC7B,iBAAiB,EAAE,MAAM,OAAO,CAAC,MAAM,CAAC,CAAC;IACzC,QAAQ,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IAC5C,aAAa,EAAE,CAAC,OAAO,EAAE,MAAM,EAAE,GAAG,EAAE,aAAa,KAAK,OAAO,CAAC,OAAO,CAAC,CAAC;IACzE,kBAAkB,EAAE,CAClB,QAAQ,EAAE,MAAM,EAChB,KAAK,EAAE,MAAM,EACb,OAAO,EAAE,yBAAyB,KAC/B,OAAO,CAAC,8BAA8B,CAAC,CAAC;IAC7C,mCAAmC,EAAE,OAAO,mCAAmC,CAAC;IAChF,0BAA0B,EAAE,OAAO,0BAA0B,CAAC;IAC9D,UAAU,EAAE,MAAM,CAAC,UAAU,CAAC;CAC/B;AA46ED,wBAAgB,mBAAmB,CACjC,SAAS,GAAE,OAAO,CAAC,yBAAyB,CAAM,GACjD,OAAO,CAsKT"}

package/dist/commands/config/index.js

@@ -6,6 +6,7 @@ import { confirmAction, } from '../shared/shared.prompts.js';
 import { readGlobalOptions } from '../shared/shared.utils.js';
 import { compileDispatchCeilingPreset } from '../../config/dispatch-ceiling-preset.js';
 import { normalizeDispatchMatrix, validateDispatchRouteTarget, walkDispatchMatrix, } from '../../config/dispatch-matrix.js';
+import { formatDispatchNotices, terminalReviewerNoticesForMatrix, } from '../../config/dispatch-notices.js';
 import { dispatchPolicyModeDescription, dispatchPolicyPolicyDescription, managedDispatchPolicyValueList, } from '../../config/dispatch-policy-options.js';
 import { VALID_DISPATCH_POLICY_MODES, VALID_MANAGED_DISPATCH_POLICIES, MAX_GATE_TIMEOUT_MS, MIN_GATE_TIMEOUT_MS, isValidGateTimeoutMs, normalizeWorkflowPostImplementSequence, readOatConfig, readOatLocalConfig, readUserConfig, writeOatConfig, writeOatLocalConfig, writeUserConfig, } from '../../config/oat-config.js';
 import { resolveEffectiveConfig, } from '../../config/resolve.js';
@@ -508,7 +509,7 @@ const CONFIG_CATALOG = [
         defaultValue: 'unset',
         mutability: 'read/write',
         owningCommand: 'oat config set workflow.hillCheckpointDefault <value>',
-        description: 'Default HiLL checkpoint behavior in oat-project-implement: "every" pauses after every phase, "final" pauses only after the last phase. When unset, the skill prompts. Resolution: env > local > shared > user > default.',
+        description: 'Default HiLL checkpoint behavior in oat-project-implement: "every" pauses after every phase, "final" pauses only after the last phase. When unset, the skill prompts. Resolution: local > shared > user > default.',
     },
     {
         key: 'workflow.archiveOnComplete',
@@ -519,7 +520,7 @@ const CONFIG_CATALOG = [
         defaultValue: 'unset',
         mutability: 'read/write',
         owningCommand: 'oat config set workflow.archiveOnComplete <true|false>',
-        description: 'Skip the "Archive after completion?" prompt in oat-project-complete. When unset, the skill prompts. Resolution: env > local > shared > user > default.',
+        description: 'Skip the "Archive after completion?" prompt in oat-project-complete. When unset, the skill prompts. Resolution: local > shared > user > default.',
     },
     {
         key: 'workflow.createPrOnComplete',
@@ -530,7 +531,7 @@ const CONFIG_CATALOG = [
         defaultValue: 'unset',
         mutability: 'read/write',
         owningCommand: 'oat config set workflow.createPrOnComplete <true|false>',
-        description: 'Skip the "Open a PR?" prompt in oat-project-complete. When true, completion auto-triggers PR creation. Resolution: env > local > shared > user > default.',
+        description: 'Skip the "Open a PR?" prompt in oat-project-complete. When true, completion auto-triggers PR creation. Resolution: local > shared > user > default.',
     },
     {
         key: 'workflow.postImplementSequence',
@@ -541,7 +542,7 @@ const CONFIG_CATALOG = [
         defaultValue: 'unset',
         mutability: 'read/write',
         owningCommand: "oat config set workflow.postImplementSequence '<legacy-or-json>'",
-        description: 'Default post-implementation chaining. Legacy strings remain supported unchanged. Structured JSON uses {"preApproval":[...],"postApproval":[...]} with the canonical sequence steps. Plain get/list/dump output serializes structured values as compact JSON; get --json preserves the object value. When unset, the skill prompts. Resolution: env > local > shared > user > default.',
+        description: 'Default post-implementation chaining. Legacy strings remain supported unchanged. Structured JSON uses {"preApproval":[...],"postApproval":[...]} with the canonical sequence steps. Plain get/list/dump output serializes structured values as compact JSON; get --json preserves the object value. When unset, the skill prompts. Resolution: local > shared > user > default.',
     },
     {
         key: 'workflow.reviewExecutionModel',
@@ -552,7 +553,7 @@ const CONFIG_CATALOG = [
         defaultValue: 'unset',
         mutability: 'read/write',
         owningCommand: 'oat config set workflow.reviewExecutionModel <value>',
-        description: 'Default execution model for the final review step in oat-project-implement: "subagent" dispatches a review subagent, "inline" runs the review in-context, "fresh-session" prints guidance for running the review in a separate session (with an escape hatch to subagent/inline). When unset, the skill prompts. Resolution: env > local > shared > user > default.',
+        description: 'Default execution model for the final review step in oat-project-implement: "subagent" dispatches a review subagent, "inline" runs the review in-context, "fresh-session" prints guidance for running the review in a separate session (with an escape hatch to subagent/inline). When unset, the skill prompts. Resolution: local > shared > user > default.',
     },
     {
         key: 'workflow.autoReviewAtHillCheckpoints',
@@ -563,7 +564,7 @@ const CONFIG_CATALOG = [
         defaultValue: 'unset',
         mutability: 'read/write',
         owningCommand: 'oat config set workflow.autoReviewAtHillCheckpoints <true|false>',
-        description: 'Automatically run the extra lifecycle review when a HiLL checkpoint is reached. This does not control Tier 1 per-phase oat-reviewer gates. When unset, the skill prompts. Resolution: env > local > shared > user > legacy autoReviewAtCheckpoints > default.',
+        description: 'Automatically run the extra lifecycle review when a HiLL checkpoint is reached. This does not control Tier 1 per-phase oat-reviewer gates. When unset, the skill prompts. Resolution: local > shared > user > legacy autoReviewAtCheckpoints > default.',
     },
     {
         key: 'workflow.autoNarrowReReviewScope',
@@ -571,10 +572,10 @@ const CONFIG_CATALOG = [
         file: '.oat/config.local.json | .oat/config.json | ~/.oat/config.json',
         scope: 'workflow',
         type: 'boolean',
-        defaultValue: 'unset',
+        defaultValue: 'true',
         mutability: 'read/write',
         owningCommand: 'oat config set workflow.autoNarrowReReviewScope <true|false>',
-        description: 'Auto-narrow re-review scope to fix-task commits in oat-project-review-provide when re-reviewing completed fix tasks. Has no effect on initial reviews (there is nothing to narrow to). When unset, the skill prompts. Resolution: env > local > shared > user > default.',
+        description: 'Automatically narrows re-review scope to changes since the prior same-lineage review. Narrowing is enabled by default; false opts out to full scope. Has no effect on initial reviews (there is nothing to narrow to). Resolution: local > shared > user > default.',
     },
     {
         key: 'workflow.autoArtifactReview.plan',
@@ -1698,45 +1699,65 @@ function applyDispatchMatrixRecommendation(workflow, recommendation) {
         },
     };
 }
+async function effectiveTerminalReviewerNotices(repoRoot, userConfigDir, dependencies) {
+    const resolved = await dependencies.resolveEffectiveConfig(repoRoot, userConfigDir, dependencies.processEnv);
+    const effectiveProviders = {};
+    for (const config of [resolved.user, resolved.shared, resolved.local]) {
+        for (const [provider, value] of Object.entries(config.workflow?.dispatchCeiling?.providers ?? {})) {
+            const existing = effectiveProviders[provider];
+            effectiveProviders[provider] =
+                isRecord(existing) && isRecord(value)
+                    ? { ...existing, ...value }
+                    : value;
+        }
+    }
+    return terminalReviewerNoticesForMatrix(effectiveProviders);
+}
 async function adoptDispatchMatrixRecommendation(repoRoot, userConfigDir, options, context, dependencies) {
     const source = options.surface === 'auto' ? 'local' : options.surface;
     const recommendation = await loadDispatchMatrixRecommendation(dependencies);
     if (source === 'user') {
         const userConfig = await dependencies.readUserConfig(userConfigDir);
         await validateRecommendationCells(repoRoot, recommendation, dependencies, context.logger.warn);
+        const workflow = applyDispatchMatrixRecommendation(userConfig.workflow, recommendation);
         await dependencies.writeUserConfig(userConfigDir, {
             ...userConfig,
-            workflow: applyDispatchMatrixRecommendation(userConfig.workflow, recommendation),
+            workflow,
         });
         return {
             key: 'workflow.dispatchCeiling.providers',
             value: recommendation.version,
             source,
+            notices: await effectiveTerminalReviewerNotices(repoRoot, userConfigDir, dependencies),
         };
     }
     if (source === 'local') {
         const localConfig = await dependencies.readOatLocalConfig(repoRoot);
         await validateRecommendationCells(repoRoot, recommendation, dependencies, context.logger.warn);
+        const workflow = applyDispatchMatrixRecommendation(localConfig.workflow, recommendation);
         await dependencies.writeOatLocalConfig(repoRoot, {
             ...localConfig,
-            workflow: applyDispatchMatrixRecommendation(localConfig.workflow, recommendation),
+            workflow,
         });
         return {
             key: 'workflow.dispatchCeiling.providers',
             value: recommendation.version,
             source,
+            notices: await effectiveTerminalReviewerNotices(repoRoot, userConfigDir, dependencies),
         };
     }
     const sharedConfig = await dependencies.readOatConfig(repoRoot);
     await validateRecommendationCells(repoRoot, recommendation, dependencies, context.logger.warn);
+    const workflow = applyDispatchMatrixRecommendation(sharedConfig.workflow, recommendation);
     await dependencies.writeOatConfig(repoRoot, {
         ...sharedConfig,
-        workflow: applyDispatchMatrixRecommendation(sharedConfig.workflow, recommendation),
+        workflow,
     });
     return {
         key: 'workflow.dispatchCeiling.providers',
         value: recommendation.version,
         source,
+        notices: await effectiveTerminalReviewerNotices(repoRoot, userConfigDir, dependencies),
     };
 }
 function formatList(values) {
@@ -1870,6 +1891,9 @@ async function runAdopt(templateArg, options, context, dependencies) {
         }
         else {
             context.logger.info(`Adopted dispatch matrix recommendation ${result.value} to ${result.source} config.`);
+            if (result.notices.length > 0) {
+                context.logger.info(formatDispatchNotices(result.notices));
+            }
         }
         process.exitCode = 0;
     }

package/dist/commands/project/dispatch-ceiling/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../../src/commands/project/dispatch-ceiling/index.ts"],"names":[],"mappings":"AAGA,OAAO,EAEL,KAAK,cAAc,EACnB,KAAK,aAAa,EACnB,MAAM,sBAAsB,CAAC;AAa9B,OAAO,EAUL,KAAK,uBAAuB,EAY7B,MAAM,oBAAoB,CAAC;AAC5B,OAAO,EAEL,KAAK,cAAc,EAEpB,MAAM,iBAAiB,CAAC;AAsBzB,OAAO,EAAE,OAAO,EAAU,MAAM,WAAW,CAAC;AAiC5C,UAAU,2BAA2B;IACnC,mBAAmB,EAAE,CAAC,OAAO,EAAE,aAAa,KAAK,cAAc,CAAC;IAChE,kBAAkB,EAAE,CAAC,GAAG,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IACrD,sBAAsB,EAAE,CACtB,QAAQ,EAAE,MAAM,EAChB,aAAa,EAAE,MAAM,EACrB,GAAG,EAAE,MAAM,CAAC,UAAU,KACnB,OAAO,CAAC,cAAc,CAAC,CAAC;IAC7B,oBAAoB,EAAE,CAAC,QAAQ,EAAE,MAAM,KAAK,OAAO,CAAC,uBAAuB,CAAC,CAAC;IAC7E,QAAQ,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IAC5C,UAAU,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,OAAO,CAAC,OAAO,CAAC,CAAC;IAC/C,UAAU,EAAE,MAAM,CAAC,UAAU,CAAC;CAC/B;AAmnFD,wBAAgB,mCAAmC,CACjD,SAAS,GAAE,OAAO,CAAC,2BAA2B,CAAM,GACnD,OAAO,CAkGT"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../../src/commands/project/dispatch-ceiling/index.ts"],"names":[],"mappings":"AAGA,OAAO,EAEL,KAAK,cAAc,EACnB,KAAK,aAAa,EACnB,MAAM,sBAAsB,CAAC;AAkB9B,OAAO,EAUL,KAAK,uBAAuB,EAY7B,MAAM,oBAAoB,CAAC;AAC5B,OAAO,EAEL,KAAK,cAAc,EAEpB,MAAM,iBAAiB,CAAC;AAyBzB,OAAO,EAAE,OAAO,EAAU,MAAM,WAAW,CAAC;AAiC5C,UAAU,2BAA2B;IACnC,mBAAmB,EAAE,CAAC,OAAO,EAAE,aAAa,KAAK,cAAc,CAAC;IAChE,kBAAkB,EAAE,CAAC,GAAG,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IACrD,sBAAsB,EAAE,CACtB,QAAQ,EAAE,MAAM,EAChB,aAAa,EAAE,MAAM,EACrB,GAAG,EAAE,MAAM,CAAC,UAAU,KACnB,OAAO,CAAC,cAAc,CAAC,CAAC;IAC7B,oBAAoB,EAAE,CAAC,QAAQ,EAAE,MAAM,KAAK,OAAO,CAAC,uBAAuB,CAAC,CAAC;IAC7E,QAAQ,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,OAAO,CAAC,MAAM,CAAC,CAAC;IAC5C,UAAU,EAAE,CAAC,IAAI,EAAE,MAAM,KAAK,OAAO,CAAC,OAAO,CAAC,CAAC;IAC/C,UAAU,EAAE,MAAM,CAAC,UAAU,CAAC;CAC/B;AAkwFD,wBAAgB,mCAAmC,CACjD,SAAS,GAAE,OAAO,CAAC,2BAA2B,CAAM,GACnD,OAAO,CA0GT"}