Refactor homebrew tasks to use loop instead of with_items - #2

Fallback-2 matched the digit in a redirection (e.g. 'gh pr merge
2>/dev/null' extracted PR #2), short-circuiting the reliable
'gh pr view' fallback and writing a wrong PR number into tasks.md.
Tighten the pattern to a bare numeric argument only.

* docs(spec): add pair-flow spec bundle

Spec-driven pipeline that pairs human and agent from comprehension
through execution and orchestration. Defines /spec-draft, /spec-kickoff,
/execute-task, /orchestrate, /resume, the new Agent-resolvable finding
bucket, and the inbox + tmux dashboard substrate.

Status: Draft. 14 tasks in dependency order; 10 open questions and 7
deferred items captured for follow-up discussion before implementation.

* docs(spec): resolve pair-flow open questions, add D-19 through D-30

Encodes the decisions from the 2026-05-22 review pass:

- D-19/D-20: two-file config model (~/.claude/pair-flow.yml tracked
  defaults + ~/.claude/pair-flow.local.yml agent-maintained), with
  discovery flow that never silently writes. Replaces the spec-config.yml
  proposal that would have leaked personal workflow into shared repos.
- D-21: PRs are always drafts; no auto-merge at any tier. Removes the
  v3 auto-merge from Deferred and moves to Out of scope as permanent.
- D-22/D-23: dashboard visual language (red/orange/blue/yellow/green/grey)
  with sort order, plus inbox heartbeat (30s tick, 2-minute stale sweep,
  24h legacy fallback).
- D-24 through D-30: bundle sizing, CI retry, path-guard scope, brief
  invalidation scope, validator reuse, PR-merge detection, telemetry
  layout.

Adds Task 3.5 for the pair-flow config helper. Resolves nine of the ten
open questions; only codex availability on personal/alt remains, gated
on Task 1's investigation.

* docs(spec): pair-flow second-pass clarifications (D-31 through D-48)

Resolves the second round of clarifications surfaced after the initial
spec was committed:

- D-31: Spec status lifecycle (Draft -> Active -> Done) with explicit
  ownership of each transition.
- D-32: Branch naming convention for orchestrator-created branches.
- D-33: No bypass mode for /spec-kickoff; manual work remains available.
- D-34: Inbox keyed per Claude Code session UUID, presented per worktree.
- D-35 through D-48: kickoff walkthrough on partial invalidation,
  orchestrator behavior on unkickoffed specs, cross-spec concurrency,
  measurement plan convention, in-session skill composition,
  Last-reviewed auto-update, incremental brief writing, spec-looks-wrong
  escalation, gh auth failure handling, worktree ownership split,
  validator status-aware enforcement, brief retention, uncommitted
  changes on /resume, and no separate audit log.

Adds Task 3.6 for the validator port + extension (a precondition for
/spec-draft and /spec-kickoff). Extends /spec-kickoff and /orchestrate
tasks with status-flipping responsibilities. Documents the spec-status
and measurement-plan conventions in specs/README.md.

The codex availability question (last remaining open question after the
first clarification pass) is also resolved: codex is available on both
work and personal profiles.

* docs(spec): pair-flow D-49 through D-51 (brief commits, validator scope, rewrite threshold)

- D-49: Kickoff briefs are committed to git as part of the spec bundle,
  not gitignored. They're decisions, not secrets.
- D-50: Validator v1 enforces structural fields only (stable ID,
  Done-when, Dependencies, Citations). Measurement plan and Last
  reviewed are recommended but not blocking.
- D-51: Wholesale-brief invalidation fires when both requirements.md
  and design.md change in the same commit, or when >50% of REQ-IDs
  or D-IDs change in a single commit. Otherwise section-scoped.

Adds REQ-A2.11 and REQ-A2.12 to anchor the requirements.

* docs(spec): pair-flow D-52, D-53 plus lock-timing clarification

Final consistency pass surfaced two gaps and two adjacent fixes:

- D-52: One task per /orchestrate invocation; intra-spec parallelism
  comes from multiple invocations (manual via parallel tmux panes, or
  the scheduled runner picking up successive ready tasks). Matches the
  user's existing "different workstreams in parallel" workflow.
- D-53: /execute-task weighs performance, security, and system-wide
  implications during research and records tradeoffs in the kickoff
  brief's risk register. CI is the hard gate; performance regressions
  that pass CI are still user-observable and excluded from
  Agent-resolvable by D-3.
- D-17 fix: reconcile lock path with D-37 (per-spec, not repo-wide) and
  pin lock-release timing (released before /execute-task runs, not held
  for its full duration).
- D-37 wording: clarify that "same-spec concurrency blocked" applies
  only to state-changing moves, not to concurrent task execution.

REQ-B1.5 expanded with explicit "industry standards, popular projects"
language and the performance/security/system-wide weighing instruction.
REQ-D2.2 added with the parallelism contract.

* docs(spec): pair-flow Task 1 — panel-* underuse diagnosis

Diagnosis: /panel-* shipped 2026-05-15, mid 30-day window. /copilot-* count
(42) all pre-release; /panel-* count (14) all post-release; zero copilot
invocations after panel landed. Not underuse, just availability.

Recommendation: keep panel as default. Confirms D-6 (codex-only default,
drop "provisional") and D-12 (panel-pairing demoted to escalation, /polish
as default convergence). One LAN-Ollama auto-mode classifier denial logged
as follow-up.

* feat(claude): pair-flow Task 3.6 — spec-validate.sh

Structural validator for spec bundles. Checks four-file presence, Status:
declaration, and per-task fields (stable ID, Done when, Dependencies,
Citations). Status-aware: Draft warns (exit 0), Active errors (exit 1).
Also flags prose-only REQs (the tecpan/specs/org counter-example pattern).

Self-tests pass on the three target fixtures:
- tecpan/specs/settings (Active): 0 errors, 0 warnings
- tecpan/specs/org (no status, prose REQs, bare tasks): 0 errors, 27 warnings
- specs/pair-flow (Draft, this bundle): 0 errors, 0 warnings

The REQ-G7.1 Gherkin scenarios verified via synthetic fixture: same gap
warns on Draft (exit 0), errors on Active (exit 1). Materializes through
the existing roles/osx/tasks/osx.yml scripts/ directory symlink; no
per-file symlink task added.

* feat(claude): pair-flow Task 3.5 — config helper + defaults file

Adds roles/osx/files/claude/pair-flow.yml (tracked defaults: codex-only panel
backend, 1h stale-lock threshold, 30s inbox heartbeat) plus a per-file
symlink task in osx.yml materializing it to ~/.claude/pair-flow.yml.

Adds roles/osx/files/claude/scripts/pair-flow-config.sh with five
subcommands: repo, defaults, repo-class, confirm-repo-class <value>, show.
Per D-20 the helper never silently writes — repo-class on an unknown repo
exits 2 with `needs-confirmation:<inferred>`; only confirm-repo-class
persists. Inference reads gh pr list and filters known bot logins
(*[bot], copilot-*, dependabot*, renovate*, github-actions*) plus
PR-author self-reviews so a solo repo with copilot review history is
not misclassified as multi-reviewer.

Smoke-tested end-to-end on this dotfiles repo. spec-validate.sh still
emits zero errors and zero warnings on specs/pair-flow.

* feat(claude): pair-flow Task 2 — file-path PreToolUse guard

Adds roles/osx/files/claude/scripts/path-guard.sh wired from settings.json
as a PreToolUse hook with matcher Read|Edit|Write (Bash and NotebookEdit
out of scope per D-26). The hook denies via the PreToolUse JSON envelope
with permissionDecision=deny and a structured reason so the model gets
useful context, not a stderr blob.

Blocks: Read/Edit on missing file or directory path; Write on missing
parent directory. Allows everything else. Smoke-tested seven cases via
direct stdin; the hook returns the correct deny envelope or silent allow
in each. End-to-end behavior needs a fresh session because PreToolUse
hooks load at session start; the materialized settings.json is updated
for the next session to pick up.

The 82/month file-path-mistake baseline (April-May 2026 friction analysis)
will be re-measured in the next 30-day window per Task 2's Measurement
plan.

* feat(claude): pair-flow Task 6 — /spec-kickoff skill

Adds roles/osx/files/claude/commands/spec-kickoff.md, the didactic
walkthrough that produces signed-off kickoff briefs. Seven sections
(Goal+glossary, Requirements, Design, Verification, Task graph, Risk
register, Sign-off) walked per a read → restate → surface implicit
terms → Socratic checks → D-42 inconsistency gate → wait-for-sign-off
→ incremental-append pattern (D-41). Retrofit mode patches tasks.md
at the Task-graph step with per-patch red-line. Partial-invalidation
re-walk uses D-27 section scope plus D-51 wholesale-rewrite triggers.
Sign-off flips all four spec files Draft → Active via Edit and bumps
Last reviewed: (REQ-A2.9, D-40); validator failure post-flip reverts
and halts. Repo-class discovery via pair-flow-config.sh with the
needs-confirmation semantics from REQ-D9.1 / D-20; never silently
writes.

Pre-flight stages traced cleanly against specs/pair-flow/: validator
0/0, repo-class solo, no prior brief. The interactive bootstrap-test
per REQ-A2.7 is the user's move and the D-7 quality gate.

* feat(claude): pair-flow Task 3 — inbox + tmux dashboard substrate

Cross-session awareness layer (REQ-F). Six inbox-*.sh scripts in
roles/osx/files/claude/scripts/ (materialized via the existing
directory symlink): write/heartbeat/sweep/session-start/status/
dashboard. Hooks wired in settings.json: SessionStart spawns the
heartbeat in the background, Stop transitions to idle, Notification
(permission_prompt | idle_prompt) transitions to awaiting-input
alongside the existing notify-event.sh chain. tmux.conf gains a
`prefix + i` popup binding for the dashboard plus a second dracula
custom plugin for the status-bar triplet (⚠N ●N ◆N).

Display-only v1 (no row-jump) and triplet status segment per user
pick on PR #28. macOS notifications reuse notify-event.sh's
tnotify-send chain rather than duplicating it. Heartbeat self-exits
when its claude PID dies (walks up to 6 ancestors looking for
claude/claude-code) or when the inbox file is removed, so no Stop
cleanup hook is required and orphan processes do not accumulate.
The sweeper combines the spec's 2-min heartbeat threshold (D-23)
with a same-host PID-liveness backstop for kill -9 leaks.

Sync mechanism (iCloud vs Syncthing) stays Awaiting input; entries
are single-writer atomic-rename sub-1KB JSON so either works.

Smoke-tested end-to-end: hook init + tick + transition + sweep all
behave; dashboard renders 6 mixed states with correct colours, sort
weights, column truncation; status segment computes correct counts;
heartbeat self-exits within 30s of its inbox file being removed.
v1 verification's two-concurrent-sessions + kill-9 + macOS-notify
checks are deferred to the user (require fresh-session hook load).

* feat(claude): pair-flow Task 8 — Agent-resolvable bucket

Add the four-bucket Finding Categorization to user-global CLAUDE.md
(Auto-applicable / Agent-resolvable / Needs sign-off / Needs human
judgment). The new Agent-resolvable bucket carries the REQ-C1.2 five-
condition predicate (failing regression test, test passes after fix,
full project CI passes, kickoff alignment, no hard disqualifier), with
a repo-class behavior split (solo auto-applies, multi-reviewer surfaces
for review with evidence) routed through ~/.claude/scripts/pair-flow-
config.sh repo-class.

/polish and /panel-pairing gain a pre-flight step that resolves repo-
class and locates the active kickoff brief, a 4-bucket categorize/
loop-fate path, and a new step (d-AR) Agent-resolvable application
flow (write failing test → confirm failure mode matches finding → fix
→ test passes → wider CI → verify kickoff alignment → record
evidence). Iteration summaries now log repo-class, brief path, all
four bucket counts, and per-item Agent-resolvable evidence. Auto-
execution invariants extended.

/peer-review updated to four tables with an Agent-resolvable thread
table and matching option semantics. /code-review's descriptive
reference updated to four-bucket. /copilot-pairing left untouched; it
inherits by reference from CLAUDE.md.

specs/pair-flow/tasks.md: Task 8 moved to Completed.

* feat(claude): pair-flow Task 9 — /polish standalone-mode PR creation

Add an "Invocation mode" section to polish.md that splits behavior on
the literal --nested flag in $ARGUMENTS. Standalone (default) pushes
and opens or updates a draft PR after a normal exit; nested mode (used
by /execute-task in the pair-flow design) hands back to the parent for
PR creation.

The new "Standalone-mode PR creation" section mirrors /self-review's
shape: push, check for an existing PR via gh pr view and reuse it,
otherwise template / convention discovery then gh pr create --draft
with a body listing the Polish notes (per-iteration Auto-applicable
and Agent-resolvable items, test paths and kickoff-brief citations,
plus any unresolved Needs sign-off / Needs human judgment items).

Auto-execution invariants split into in-loop (no push, no PR ever)
and post-loop (standalone may push and open a draft PR, nested may
not). Force-push, amend, squash, and rebase remain forbidden at all
times. Safety stops suppress PR creation in both modes.

specs/pair-flow/tasks.md: Task 9 moved to Completed.

* feat(claude): pair-flow Task 4 — tasks.md state conventions + auto-update hook

Document the canonical tasks.md state sections (Completed, In progress,
Awaiting input, Deferred, Out of scope) and the In progress annotation
format in specs/README.md. Phase values: implementing, polish iter N,
PR #M draft. Bullet shape for Completed pinned.

Add roles/osx/files/claude/scripts/tasks-pr-sync.sh wired from settings
.json under hooks.PostToolUse with matcher "Bash". The script filters
`gh pr (create|merge)` from the tool call, parses the D-32 branch
(pair-flow/<spec>/task-<ids>, including bundles and dotted IDs),
extracts the PR number from tool_response.stdout via URL match, then
`#N` literal, then argv, then `gh pr view` fallback, and rewrites
specs/<spec>/tasks.md.

On open: move task block(s) Forward plan → In progress, insert the
two annotation lines right after the H3; strip existing Status /
Last-activity bullets first so the hook is idempotent. On merge:
replace the block with a one-line Completed bullet referencing the
PR. Bundle order preserved. Silent on no-op cases; python failures
log to stderr and leave tasks.md untouched.

Smoke-tested seven scenarios via synthetic stdin against ephemeral
fixtures (single-task open, gh pr merge with realistic #N stdout,
bundle open with correct order, non-pair-flow branch no-op, non-Bash
tool no-op, idempotent re-open). Out-of-session merge reconciliation
remains Task 12's scheduled-runner job.

* feat(claude): pair-flow Task 5 — /resume context-load skill

* feat(claude): pair-flow Task 10 — /execute-task skill

* feat(claude): pair-flow Task 11 — /orchestrate v1 skill

* feat(claude): pair-flow Task 12 — scheduled runner documentation

* feat(claude): pair-flow Task 7 — /spec-draft skill

* docs(spec): pair-flow kickoff-brief; status Draft -> Active

* feat(claude): pair-flow Task 13 — spec amendments, contract checker, observation convention

Apply kickoff-brief amendments: D-6 profile-aware panel backends (codex
on work, gemini on personal/alt), D-17 stale-lock threshold reduced from
1h to 15m, D-24 effort-based fallback for low-data repos.

Add skill-contracts.sh contract-consistency checker wired as a lefthook
pre-commit job. Checks cross-file invariants (D-32 branch naming,
four-bucket presentation, --nested flag, pair-flow-config.sh, D-21
always-draft, D-33 Active gate, D-36 kickoff gate) across skill files.

Add observation convention to /execute-task and /polish (Risk 6 Layers
1/1b: passive record of opportunities for future /spec-draft seeds).

* chore(polish): iter 1, expand pair-flow-config.sh invariant coverage

Add execute-task.md and peer-review.md to the pair-flow-config.sh
check list in skill-contracts.sh. Both make load-bearing calls to
pair-flow-config.sh repo-class in their pre-flight.

* docs(spec): update Task 13 status to PR #28 draft

* feat(claude): interaction style for spec skills, composability principle, Task 15

- Add interaction style sections to /spec-draft and /spec-kickoff:
  progress indicator, progressive disclosure, visual aids, selectors
  over open-ended questions, smart defaults, running summary
- Add composability-by-default design principle to user-global CLAUDE.md
- Add Task 15 (work project e2e validation) to pair-flow tasks
- Add standalone project extraction to Deferred with gate conditions

* feat(claude): polish handoff presentation with clustered decisions

Add explicit handoff presentation rules to /polish post-loop flow.
Items sharing a decision axis (e.g., test coverage gaps) cluster into
one question instead of individual confirmations. Reduces human
interaction count without losing granularity.

* feat(claude): panel-pairing handoff presentation with clustered decisions

Same fix as polish: when the autonomous loop hands off residual items,
cluster shared-axis findings and use batched decisions instead of
defaulting to one-by-one confirmation.

* fix(claude): orchestrate worktree context switch via EnterWorktree

/orchestrate said "navigate to the worktree" but didn't specify how,
so /execute-task ran in the wrong directory. Now explicitly uses
EnterWorktree/ExitWorktree. Also documents /loop wrapping for
multi-task autopilot.

* docs(spec): defer four-file format meta-spec until extraction gate

Gated on the standalone extraction goal — current sources
(specs/README.md, validator, /spec-draft) are sufficient for personal
use, and a meta-spec would add a sync burden without immediate value.

* docs(spec): defer Dynamic Workflows integration investigation

Two integration surfaces identified (discovery rigor fan-out across
review skills, /execute-task conditional fan-out for multi-file
mechanical work). Gated on Anthropic publishing the API.

* fix(claude): self-review base diff uses origin/main, not stale local

A stale local main in a long-lived worktree inflates the diff with
already-merged commits. Fetch first and diff against the remote-tracking
base; fall back to local main only when no remote is configured.

* fix(claude): four findings tables in self-review and panel-review

CLAUDE.md Finding Categorization defines four buckets (added
Agent-resolvable), but /self-review and /panel-review still said
'three findings tables'. Aligned to the canonical four.

* docs(spec): pair-flow v1 retrospective

Task 13 deliverable. Captures e2e validation findings across dotfiles,
STEAI-748 work project, and tecpan kickoffs. Sections: what worked,
what needed iteration during the run, concrete tunings per Task 13's
Done-when, personal-preference vs generalizable for the
standalone-extraction gate, and what v1 did not validate.

Recommends holding the extraction gate until at least one tecpan
/orchestrate -> draft PR cycle closes the orchestration validation gap.

* feat(claude): gemini panel backend on personal/alt per D-6

- Brewfile: install gemini-cli
- scripts/claude-gemini-auth-sync.sh: read GEMINI_API_KEY from 1Password
  (item hvscsuq25owvgrqt235xwlfmgy), write ~/.gemini/.api-key mode 0600
- roles/fish/files/gemini.fish + symlink task: export GEMINI_API_KEY
  from that file into the shell environment
- roles/osx/tasks/homebrew.yml: call the sync script on non-work hosts
  (skipped on CI and inventory_hostname == 'work')
- pair-flow-config.sh: merge top-level keys from pair-flow.local.yml
  (not just repo-class), so panel-backends: [gemini] on personal/alt
  resolves correctly per D-6 amendment
- panel-review.md + panel-pairing.md: add gemini to supported backends,
  with setup check (auth via env var) and invocation pattern
  (gemini -p "<prompt>" -o text); update default tables to D-6
  (work=codex, personal/alt=gemini)
- CLAUDE.md: align /panel-review and /panel-pairing default tables

* fix(claude): polish + panel-pairing kickoff-brief resolution with multiple Active specs

Previous heuristic ('single Active spec or D-32 branch pattern, else
disable Agent-resolvable') gave up too early when multiple specs are
Active. Walk five steps in order: D-32 branch -> single Active spec ->
branch-name token match -> diff-scope >=80% under one spec -> ask the
user. Only disable the bucket when no Active spec exists at all.

* docs(spec): close out Task 13 (tecpan /orchestrate PR #161 merged)

tecpan PR #161 (settings Tasks 3, 4, 5 via /orchestrate, merged
2026-06-04) is the missing piece for Task 13's 'at least one tecpan
task shipped via the full pipeline' Done-when. Moves Task 13 to
Completed, updates the retrospective to reflect the orchestration
validation evidence, and flips the extraction-gate recommendation to
'gate has enough evidence to fire.'

* feat(claude): orchestrate worktree reuse + claude --worktree interop (D-54); autonomy-pipeline docs (Task 14)

* docs(spec): re-sign kickoff brief Section 3 for D-54 worktree reuse + D-44 amendment

* chore(polish): iter 1, fix shellcheck SC2004/SC1010 in inbox scripts

* fix(claude): guard inbox writes against empty jq output (N1) and skip malformed dashboard blob parts (N2)

* fix(claude): escape notify JSON + host-scope inbox stale sweep

- inbox-write.sh: build notify-event payloads via jq --arg so a branch or
  path containing a double-quote can't emit malformed JSON (N1).
- inbox-sweep.sh: host-scope the heartbeat-staleness/legacy deletion so a
  synced inbox can't have one host sweep another host's live entry by local
  clock when sync lag exceeds the stale threshold (J2).
- inbox-write.sh: document the accepted tick/hook-event read-modify-write
  race (self-heals within one heartbeat; flock absent on macOS) (J1).

* docs(spec): pin 5 unpinned REQs in test-spec; de-em-dash CLAUDE.md skill list

- test-spec.md: add verification entries for REQ-A2.9/A2.11/A2.12/B1.10/B1.11
  so every REQ is pinned per REQ-A1.5 (N2).
- CLAUDE.md: replace em-dashes with colons in the pipeline skill list to
  honor the repo's own Writing Style rule (N3).

* docs(claude): gate /polish observations on pair-flow adoption

/polish can run in any repo; only write specs/_observations/opportunities.md
when the repo has adopted pair-flow (a specs/<feature>/requirements.md exists).
/execute-task always runs against an Active spec, so it notes the premise is
guaranteed there.

* docs(claude): scope panel Backend-failure stop to non-recovered failures

Clarify in /panel-review and /panel-pairing that the Backend failure stop
condition fires only on a non-recovered failure (final non-zero exit, empty or
unparseable output, auth lost with no successful retry), not on transient
quota/rate-limit/retry notices the CLI prints while retrying internally and
then still returning a valid result. Judge by final outcome, not stderr.

* docs(observations): seed note to decouple review tooling from pair-flow

Capture scope for a future separate branch: panel-backends config mis-homed in
pair-flow.yml, employer org names leaked in panel-review.md:23, CLAUDE.md
over-couples the review workflows. Includes the leak surface map and the
work-detection mechanism decision.

* fix(claude): tasks-pr-sync PR-number regex ignores shell redirections

Fallback-2 matched the digit in a redirection (e.g. 'gh pr merge
2>/dev/null' extracted PR #2), short-circuiting the reliable
'gh pr view' fallback and writing a wrong PR number into tasks.md.
Tighten the pattern to a bare numeric argument only.

* docs(spec): sync pair-flow status and stale-lock threshold across docs

- specs/README.md: pair-flow row Draft -> Active; cold-start points at
  kickoff-brief + tasks.md state (was stale 'begins at Task 1/2').
- requirements.md: drop stale 'Implementation begins at Task 1/2'.
- design.md D-9: 'webhook-triggered hook' contradicted D-29 (no webhook);
  reword to PostToolUse (in-session) or scheduled runner (out-of-session).
- tasks.md: stale-lock threshold 1h -> 15m to match pair-flow.yml/D-17;
  annotate Task 3.5 record with the D-52 reduction.

* chore(copilot): iter 1, address portability and PAIR_FLOW_HOST findings

- spec-validate.sh: replace non-POSIX \b (grep) and \y (BSD awk) regex
  boundaries with portable equivalents; \y silently failed on BSD awk,
  causing a latent false 'lacks stable numeric ID' on EOL-number headings.
- tasks-pr-sync.sh: match gh pr create/merge as an anchored command
  invocation, not a raw substring, so unrelated commands mentioning the
  string no longer rewrite tasks.md.
- inbox-heartbeat/sweep/dashboard.sh: honor PAIR_FLOW_HOST like
  inbox-write.sh's host_name(), so an overridden host name no longer
  desyncs the entry filename (heartbeat never ticking, entries mislabeled
  remote, host-scoped sweep skipped).

* chore(copilot): iter 2, portability, dep guard, and doc-vs-code alignment

- skill-contracts.sh: use grep -E for alternations instead of non-POSIX \|.
- pair-flow-config.sh: guard python3 + PyYAML before YAML read/write so a
  fresh machine fails with an actionable message, not ModuleNotFoundError.
- requirements.md REQ-F1.2: filename contract is {host}-{session}.json with
  repo/branch as JSON fields, matching the implementation.
- inbox-dashboard.sh: clarify the aggregation key is host|worktree|branch.
- panel-review.md / panel-pairing.md: reference pair-flow-config.sh show as
  the canonical config merger for panel-backends.
- homebrew.yml: correct the Gemini-sync comment (when: gates non-work+non-CI;
  the script itself requires an authenticated 1Password session).

* chore(copilot): iter 3, redact employer data + dashboard/sweep/perms fixes

Confidentiality (public repo): redact employer org names and work-project
details from files new/changed in this branch.
- panel-review.md: profile detection reads PANEL_REVIEW_PROFILE env var
  instead of a hardcoded work-org allowlist.
- peer-review.md: genericize a work-repo war-story reference.
- v1-retrospective.md: generalize the work-project evidence (ticket ids,
  teammate name, internal identifiers, production figures, repo names).
- decouple-review-tooling.md, design.md, requirements.md, test-spec.md,
  tasks.md: replace employer/work-project names with generic descriptions.
Org names also exist in main history and out-of-diff files (tm.fish,
copilot-*.md, metrics-baseline); the user handles those separately.

Findings:
- inbox-dashboard.sh: document weight 1 as error/blocked (a session dashboard
  has no stale-lock state); skip fs_epoch=0 parts so malformed entries don't
  dominate the sort.
- inbox-sweep.sh: list filters stale entries from display regardless of host
  (D-23); deletion stays host-scoped.
- inbox-write.sh / inbox-sweep.sh: best-effort chmod 700 on the inbox dir.

* chore(copilot): iter 4, fix spec-bundle contract drift + stale comment

- requirements.md REQ-G3.1: profile-aware default (work: codex; personal/alt:
  gemini) per D-6, replacing the stale 'codex on all profiles' text that
  contradicted the rest of the bundle.
- kickoff-brief.md Task Graph Reconstruction: mark it a sign-off-time snapshot
  (tasks.md is canonical) and correct the 15->16 task count.
- claude-gemini-auth-sync.sh: the default ITEM_UUID is the real 1Password item,
  not a placeholder; fix the stale comment (no behavior change).

* chore(copilot): iter 5, explicit jq dependency guards in hooks

Add an early 'command -v jq || exit 0' to path-guard.sh, inbox-session-start.sh,
and inbox-write.sh so they degrade cleanly on a machine without jq, consistent
with tasks-pr-sync.sh. The existing 2>/dev/null already suppresses the shell's
command-not-found under bash, but inbox-write.sh has write-side jq calls that
are not 2>/dev/null-guarded, and the explicit check makes the degrade-to-safe
behavior intentional rather than incidental. Verified: with jq absent all three
exit 0 silently; with jq present, path-guard still denies and inbox tick still
writes.

* chore(copilot): iter 6, tmux status wrapper guard + repo-class comment fix

- roles/tmux/files/scripts/inbox-status.sh: guard the exec so the dracula
  status slot degrades to 'inbox?' if the Claude script is not yet
  materialized (tmux role applied before osx role), instead of erroring.
- pair-flow-config.sh: fix the misleading infer_repo_class comment. The no-gh
  fallback keeps suggesting solo (the common case), but the comment no longer
  claims that 'surfaces for review' (it doesn't; solo enables auto-apply). The
  value is only a needs-confirmation suggestion the human confirms (exit 2).

(A third Copilot thread flagging the gh_pr regex '\$' as a bug was a false
positive: bash double-quote processing converts it to the '$' end anchor
before grep sees it; bare 'gh pr merge' is detected. Verified, dismissed.)

* chore(copilot): iter 7, fix inbox notify payload + orchestrate flag parsing

- inbox-write.sh maybe_notify: pass the real cwd to notify-event.sh instead of
  a pre-built 'project · branch' string. notify-event derives project+branch
  itself (basename + git -C) and prepends that context; the old code broke
  branch detection (a non-path cwd failed the git lookup in the draft-pr-ready
  case) and duplicated the context in the awaiting-input permission body.
  Verified: done -> {cwd: <real path>}, permission -> {cwd, message:'awaiting
  input'}.
- orchestrate.md: clarify that flags (e.g. --bookkeeping) are ignored when
  resolving the spec path, so '/orchestrate --bookkeeping specs/<spec>' is not
  misread as treating --bookkeeping as the path.

* chore(copilot): iter 8, refresh branch on update + jq guards + README wording

- inbox-write.sh: the existing-entry update path now refreshes .branch (guarded
  to skip empty, e.g. detached HEAD), so switching branches in the same session
  no longer leaves a stale branch in the dashboard/status forever.
- inbox-sweep.sh: add the early 'command -v jq || exit 0' guard (degrade to
  no-sweep/empty-list), consistent with the other inbox scripts.
- inbox-dashboard.sh: same jq guard (message + exit 1, matching its missing-
  SWEEPER handling) so a fresh host degrades cleanly. (Proactively drained the
  same class Copilot flagged on inbox-sweep.sh.)
- specs/README.md: clarify that status controls task-structure-check severity
  only; the four-file presence check always errors regardless of status.
Verified: stale branch refreshes on next hook-event; sweep/dashboard degrade
cleanly with jq absent.

* chore(copilot): iter 9, harden ~/.gemini perms + tmux fallback throttle

- ~/.gemini directory perms: add an Ansible 'file' task (state: directory,
  mode: 0700) in homebrew.yml before the sync task, mirroring the
  ~/.claude/inbox precedent in osx.yml, plus a best-effort 'chmod 700' in
  claude-gemini-auth-sync.sh as defense-in-depth. The dir holds the API key
  file (only the file was chmod 600 before; the dir used the umask).
- roles/tmux/files/scripts/inbox-status.sh: the iter-6 placeholder fallback
  returned immediately, which lets dracula re-invoke the custom: slot in a
  tight loop. Sleep for status-interval (default 5s) in the fallback too,
  matching the real inbox-status.sh and gpu-usage.sh siblings.

* chore(copilot): iter 10, correct host_name .local comment

The comment claimed host_name() strips a .local suffix, but only 'hostname -s'
does that (the bare-hostname fallback, used only if -s fails, may keep the
suffix). Reword the comment to describe the actual behavior; no behavior change,
and the derivation stays consistent with the other inbox scripts.

* chore(copilot): iter 10, de-duplicate awaiting-input notification

Per the chosen resolution: keep the rich direct notify-event.sh permission/idle
hooks in settings.json (they carry the actual permission message) and stop
inbox-write.sh from emitting its own generic 'awaiting input' notification on
the same transition. inbox-write still records the awaiting-input state for the
dashboard, and still owns the draft-pr-ready notification (which has no direct
hook). Removes the duplicate macOS notification on permission_prompt/idle_prompt.
Verified: awaiting-input updates state with no notification; draft-pr-ready
still fires 'done'.