docs: codex-efficiency fix-cycle spec and plan (campaign record)

docs/superpowers/specs/2026-07-30-codex-efficiency-fixes-design.md

@@ -0,0 +1,252 @@
+# Codex Efficiency Fixes — Design
+
+Date: 2026-07-30
+Status: approved by Jesse (in-session)
+Branch: `codex-efficiency-fixes` off `dev`
+
+## Sources
+
+- Eval campaign closeout: `superpowers-autoresearch/reports/2026-07-codex-efficiency-campaign.md`
+  (treatment table §4; every treatment below has a scorer and a measured
+  `dev` baseline).
+- Codex source recon: `superpowers-autoresearch/docs/2026-07-29-codex-multiagent-v2-capabilities.md`
+  (file:line citations against the Codex CLI source; grounds T2, T3, T5).
+- Published experiment write-ups: `superpowers-evals/docs/experiments/`.
+- Drew's spinout stack (PRs #2036, #2035) is **evidence, not adopted text**:
+  Jesse wants to dig into those fixes in more detail before adopting any
+  of them; they inform the problem statements only.
+
+## Goal
+
+Ship the five evidence-strong treatments from the codex-efficiency eval
+campaign as superpowers skill/doc changes, each graded against its
+pre-registered criterion by the campaign's scorers before its PR is cut.
+Phase 2 (everything else in the closeout treatment table) follows, each
+item gated on new baseline work first.
+
+## Scope decisions (settled with Jesse)
+
+- **Phase 1 = the evidence-strong five** (T1–T5 below). Phase 2 items
+  each need a failing baseline before any fix ships (discrimination
+  rule: inconclusive-by-zero is a stop).
+- **One branch, PR per treatment.** Development and batteries happen on
+  `codex-efficiency-fixes`; when a treatment beats its criterion, it is
+  cut into its own PR against `dev` with its eval evidence. No merge
+  without Jesse's per-PR approval.
+- **T4 ships cross-harness with a global regression battery** (Claude
+  Code, Codex, Gemini), variant C shape: ceremony scales, approval never
+  does.
+
+## The five treatments
+
+### T1. SDD worker-review prohibition
+
+**Evidence:** 9/9 depth-2 spawns across 4 corpora were implementer-issued
+reviewers; all 9 were same-task duplicates of the review the controller
+dispatches anyway. The dispatch contract never says review is not the
+worker's job; "self-review" in the implementer prompt gets reified into a
+reviewer subagent on harnesses where children can spawn (Codex).
+
+**Changes:**
+- `skills/subagent-driven-development/implementer-prompt.md`: an explicit
+  "You do not dispatch subagents" clause — self-review means reading your
+  own diff; the controller owns all review dispatch; a reviewer you spawn
+  duplicates a review the process already provides.
+- `skills/subagent-driven-development/SKILL.md`: one dispatch-contract
+  line in the task loop, plus a Red Flags row: "An independent review
+  would strengthen my report" → review is the controller's next step;
+  your reviewer is a duplicate seat.
+- Harness-agnostic wording (no-op where children cannot spawn).
+
+**Graded by:** `score_e6.py` (depth-2 spawns by spawner role, duplicate
+review families); `score_e5.py` for the same-scope variant.
+**Baseline:** 9/9 worker-issued, 0 counter-examples.
+**Criterion:** 0 worker-issued depth-2 spawns AND review coverage
+preserved (every task still gets exactly one controller-dispatched task
+review).
+
+### T2. Event-driven waiting
+
+**Evidence:** 60–78% of `wait_agent` calls time out in every corpus
+(dev 67.1%, spinout 60.2%). Source recon: V2 waits are event
+subscriptions, not polls — one long wait has the same wake latency as a
+10s poll at ~1/90th the calls; a completed child's FINAL_ANSWER is pushed
+into the parent's mailbox and drained into the next model request with no
+wait at all.
+
+**Changes** (`skills/using-superpowers/references/codex-tools.md`):
+- Never short-timeout poll.
+- While local work remains, do not wait — child results arrive with your
+  next turn via the mailbox.
+- When genuinely idle, issue ONE `wait_agent` with a long `timeout_ms`
+  (900000+; harness max 3600000).
+- V2 caveat stated: completion mail carries `trigger_turn=false` and will
+  not wake an idle controller — that is the one job `wait_agent` has.
+
+**Graded by:** `score_e7.py` (timeout rate, inter-poll cadence,
+cache-rebill estimate — the rebill figure stays labeled as an estimate).
+**Baseline:** dev 67.1% timeout rate.
+**Criterion:** timeout rate < 25% with no loss of task completion.
+
+### T3. codex-tools.md corrections
+
+**Evidence:** five claims in the current guidance are contradicted by the
+Codex source (all file:line-cited in the capabilities doc):
+1. `close_agent` does not exist in multi-agent V2 (V1-only). V2 LRU-evicts
+   finished children automatically; not closing costs nothing;
+   `followup_task` transparently reloads an evicted child.
+2. Fix rounds can always resume the implementer via `followup_task` —
+   dev's "if your harness cannot send another message to a spawned agent,
+   dispatch each fix round as a fresh implementer" branch is dead on V2.
+3. Role files (`~/.codex/agents/**.toml`) DO attach to spawns via
+   `agent_type` on isolated forks (0.145+).
+4. Full-history forks accept `model`/`reasoning_effort` overrides; only
+   `agent_type` is refused. (Isolated forks remain the SDD guidance for
+   context-hygiene reasons, stated accurately.)
+5. Dispatch guidance must never name non-V2 model presets — the V2 spawn
+   allowlist is v2 presets only; others hard-error.
+
+**Changes:** rewrite the multi-agent paragraph of
+`skills/using-superpowers/references/codex-tools.md` to be
+version-honest (V1 vs V2 behavior labeled where they differ).
+
+**Graded by:** source citation (already verified); no scorer regressions
+on the shared battery. `score_e8.py` is retained as a V1/V2 schema
+detector, not a hygiene grader — no `close_agent` checklist ships.
+
+### T4. Brainstorming three-path router (variant C: approval always)
+
+**Evidence:** micro — the current HARD-GATE text pushes a bounded task to
+FULL ceremony 5/5, while Z-null (no guidance) and a three-path router
+both differentiate 5/5: the absolute wording suppresses discrimination
+the model draws natively. FULL battery — ceremony volume scales
+moderately (16.7 vs 24.0 tool calls, bounded vs arch), but the
+two-document ritual (spec file → plan file) ran unconditionally in every
+rep. The measured waste is the unconditional artifact ritual, not the
+approval gate.
+
+**Design (variant C):** three paths scale the ARTIFACT; every path keeps
+human approval before implementation:
+- **Spike** (feasibility question, explicitly throwaway): present the
+  question and the intended probe in 2–3 sentences, get a nod, go. No
+  docs. Findings return as a recommendation; anything built stays labeled
+  throwaway.
+- **Bounded** (well-scoped change to an existing, understood flow):
+  present a short design in chat, get approval, implement. No spec file,
+  no writing-plans invocation.
+- **Architectural** (restructures components, new subsystem, public
+  interface change): the full current flow — spec doc, review,
+  writing-plans.
+
+**Guards (all ship with the router):**
+- Classification is said out loud ("this looks bounded, so I'll present a
+  short design here rather than write a spec") so the human can override.
+- When in doubt between two paths, take the heavier one.
+- One-way ratchet: hidden complexity discovered mid-path upgrades the
+  path; never downgrade mid-task.
+- New Red Flags rows targeting classification-as-escape-hatch ("I'll call
+  it bounded to skip the doc").
+
+**Changes** (`skills/brainstorming/SKILL.md`): HARD-GATE keeps "no
+implementation before approval" and drops "regardless of perceived
+simplicity" as the ceremony driver; anti-pattern section reframed (the
+sin is skipping approval, not skipping documents); checklist steps 6–9
+become the architectural path; process-flow graph gains the router; Red
+Flags rows added. This is carefully-tuned content — the edit follows
+writing-skills methodology and ships only with the full eval evidence
+below.
+
+**Graded by (three layers):**
+1. **Micro** (`ceremony-path-micro.py`, adapted): variant C literal text,
+   plus adversarially ambiguous briefs the campaign never tested (a task
+   that pattern-matches bounded but hides a public interface change).
+   Criteria: spike/bounded/arch differentiate (≥4/5 per cell); ambiguous
+   briefs escalate to FULL (≥4/5); arch never downgrades (5/5).
+2. **Codex ceremony battery:** `cx-ceremony-{spike,bounded,arch}` on the
+   fix arm, 3 reps each, `score_e4.py` census. Criteria: bounded reps
+   show an approval turn but zero committed spec files and zero
+   writing-plans ritual; arch reps keep the full two-doc flow; spike reps
+   stay minimal.
+3. **Global regression battery:** the same three ceremony scenarios on
+   Claude Code and Gemini (rig work: those scenarios are currently
+   codex-gated), 3 reps each; plus the triggering acceptance check
+   ("Let's make a react todo list" auto-triggers brainstorming into the
+   full/architectural path) on all three harnesses.
+
+### T5. Explicit model on child-issued spawns
+
+**Evidence:** root spawns are 100% explicit-model at CLI 0.146 (dev
+14/14); the live gap is depth-2 — 2/2 child-issued spawns omitted
+`model`. Source recon: `model` without `reasoning_effort` resets effort
+to the MODEL's default, not the parent's.
+
+**Changes** (`skills/using-superpowers/references/codex-tools.md`):
+- Every spawn you issue — including as a child — sets `model` AND
+  `reasoning_effort`; the effort-reset trap is named.
+- Advise `[agents].default_subagent_model` and
+  `[agents].default_subagent_reasoning_effort` in `~/.codex/config.toml`
+  as the machine-level backstop for anything that slips through.
+
+**Graded by:** `score_e1.py` (per-spawn explicit-model rate, by depth) on
+the shared battery.
+**Baseline:** depth-2: 0/2 explicit.
+**Criterion:** every spawn at every depth carries explicit model +
+effort. Pre-registered caveat: if T1 eliminates depth-2 spawns entirely,
+T5 grades as root-spawn regression (hold 100%) plus doc correctness and
+is recorded inconclusive-by-zero at depth-2 — the config backstop is then
+the operative mechanism.
+
+## Grading plan
+
+- **Shared SDD battery** carries T1, T2, T5: `cx-sdd-small`, fix-branch
+  arm (`/tmp/sp-arm-fix`), 8 reps across both container lanes. Dev
+  baselines are already measured; no baseline re-runs.
+- **T4 batteries** as listed above (micro + codex ceremony + global
+  regression).
+- **Pre-registration:** every battery gets a hypothesis-log entry
+  (prediction, scorer, criterion) in
+  `superpowers-autoresearch/logs/2026-07-30-codex-efficiency-fixes.md`
+  BEFORE it runs. Standing rules carry over: append-only log, manual
+  inspection of scorer matches on fix-arm runs (non-circular
+  verification), no raw rollouts committed, correctness rides beside
+  cost in every verdict.
+- **Attribution:** orthogonal scorers on one combined branch; unexpected
+  regressions bisect by treatment commit.
+- **Budget:** shared battery ~$40, codex ceremony ~$40, global
+  regression ~$40–80, micros ~$5 → phase 1 ≈ $150–200 of the ~$850
+  remaining from the campaign's $1000.
+
+## Process
+
+- Work happens in the `codex-efficiency-fixes` worktree (branched off
+  `dev`); execution via subagent-driven-development from a written plan.
+- Skill-text changes follow writing-skills methodology.
+- Scenario/rig changes (un-gating ceremony scenarios for Claude
+  Code/Gemini, adversarial micro briefs) land in `superpowers-evals`
+  main, as authorized.
+- PR-per-treatment against `dev`, each with its eval evidence and the
+  standard identification block; merges only on Jesse's per-PR approval.
+
+## Phase 2 queue (baseline-first; not in this plan's tasks)
+
+Each item requires a failing baseline before any fix ships:
+1. **Dispatch routing / long-session drift** — needs a long-session
+   elicitation rig (fresh sessions don't reproduce the pathology at CLI
+   0.146). Drew's stack informs the treatment shape.
+2. **Verification leases / evidence receipts** — needs the
+   substring-aware duplicate counter added to `score_e3.py` first
+   (current baseline 1/23 exact-string pairs is too weak).
+3. **Remediation cap** — small-n baseline (2/3 reps) needs more reps.
+4. **Cross-task-race probe redesign** — `score_e5.py`'s probe is
+   inconclusive-by-zero by design tradeoff; needs a stronger probe.
+5. **E5 D4 shell-command parser** — fix-review-scope classifier cannot
+   parse compound commands; scorer work, not skill work.
+
+## Out of scope
+
+- Adopting Drew's spinout stack (#2036/#2035) or its text.
+- RoboRev, Codex token telemetry (separate codebases).
+- A `close_agent` hygiene checklist (V2 has no such tool — closed as
+  do-not-ship in the campaign).
+- Claude Code/Gemini-specific efficiency treatments beyond the T4
+  regression battery.

skills/subagent-driven-development/SKILL.md

@@ -197,17 +197,6 @@ Everything you paste into a dispatch prompt — and everything a subagent
 prints back — stays resident in your context for the rest of the session
 and is re-read on every later turn. Hand artifacts over as files.
 
-**Waiting on dispatched subagents:** never poll a wait interface with
-short timeouts, and never sit in one silent, open-ended wait either.
-While you have local work — ledger updates, packaging the next review,
-reading reports — keep working; child results arrive on their own.
-When you are genuinely idle, wait in bounded stretches (five to ten
-minutes, where your platform allows), and between stretches post one
-line of status and reconcile your live children: list them, and chase
-any that finished without reporting. A bounded stretch keeps nearly
-all of a long wait's efficiency while guaranteeing a stuck or lost
-child is noticed within minutes, not at the end of the session.
-
 ### 1. Dispatch the implementer
 
 Record BASE (`git rev-parse HEAD`) before dispatching — the review package

skills/using-superpowers/references/codex-tools.md

@@ -7,76 +7,7 @@ Add to your Codex config (`~/.codex/config.toml`):
 multi_agent = true
 ```
 
-This enables the multi-agent tools that skills like
-`dispatching-parallel-agents` and `subagent-driven-development` use.
-Which tools you get depends on the multi-agent version your model
-preset selects (current presets run V2; older ones run V1). Trust your
-actual tool list over any table — including this one — when they
-disagree.
-
-- **Spawning:** give children a clean context with
-  `spawn_agent {fork_turns: "none"}`; the default `"all"` copies your
-  entire transcript into the child. On Codex 0.145+, role files under
-  `~/.codex/agents/` attach to isolated forks via `agent_type`.
-  Full-history forks accept `model` and `reasoning_effort` overrides
-  (only `agent_type` is refused there) — isolated forks are the SDD
-  default for context hygiene, not because overrides require them.
-- **Fix rounds:** resume the implementer with `followup_task` — it
-  delivers your message, triggers a turn, and transparently reloads a
-  child the harness evicted. Never dispatch a fresh implementer on the
-  theory that a spawned agent cannot be messaged again; on V2 it
-  always can.
-- **Lifecycle:** V2 has no `close_agent`. Finished children are
-  evicted automatically when slots are needed; leaving them unclosed
-  costs nothing. Only V1 sessions have `close_agent` — there, close
-  reviewers when their review returns, and close each implementer
-  after its task's review passes.
-- **Model names:** never copy a model name from a skill, table, or old
-  session into `spawn_agent` without checking it against your current
-  spawn allowlist — V2 accepts only V2-capable presets and hard-errors
-  on the rest.
-
-## Waiting on children
-
-`wait_agent` is an event subscription, not a poll: a long wait wakes
-the moment a child produces mailbox activity, with the same latency as
-a short one. Short-timeout polling buys nothing and costs a tool call —
-and a context rebill — per poll. In measured sessions, roughly
-two-thirds of all wait calls were short polls that timed out.
-
-- While you still have local work, do not wait at all. A completed
-  child's final answer is pushed into your mailbox and arrives with
-  your next turn.
-- When you are genuinely idle with children outstanding, wait in
-  bounded stretches: `wait_agent` with `timeout_ms` 300000-600000
-  (5-10 minutes). After each stretch — wake or timeout — post one
-  status line, run `list_agents`, and chase any child that finished
-  without reporting. Never stack polls shorter than five minutes; the
-  event subscription wakes a bounded stretch just as fast as a short
-  one.
-- Completion mail cannot wake an idle controller (it is delivered
-  without triggering a turn); covering that idle window is
-  `wait_agent`'s only job. A stretch that times out with no activity
-  is your cue to reconcile, not to shorten the next stretch.
-
-## Model routing on spawns
-
-Every `spawn_agent` you issue — including when you are yourself a
-spawned child running a fan-out — sets `model` AND `reasoning_effort`
-explicitly, per the Model Selection rules of the skill you are
-executing. Setting `model` alone is a trap: the child's effort
-silently resets to that model's default, not to yours.
-
-Ask your human partner to add a machine-level backstop to
-`~/.codex/config.toml` so any spawn that slips through still routes to
-a deliberate tier instead of silently inheriting the session's most
-expensive model:
-
-```toml
-[agents]
-default_subagent_model = "<a mid-tier model from your spawn allowlist>"
-default_subagent_reasoning_effort = "medium"
-```
+This enables `spawn_agent`, `wait_agent`, and `close_agent` for skills like `dispatching-parallel-agents` and `subagent-driven-development`. When using subagent-driven-development, close reviewer subagents when their review returns. Keep each implementer subagent open until its task's review passes — the fix loop resumes the implementer — then close it. If your harness cannot send another message to a spawned agent, dispatch each fix round as a fresh implementer carrying the brief, the report file, and the findings.
 
 ## Environment Detection