oh-my-opencode 4.16.3 → 4.17.1

package/packages/omo-codex/plugin/components/ultrawork/directive.md

@@ -6,8 +6,7 @@
 [CODE RED] Maximum precision. Outcome-first. Evidence-driven.
 
 # Role
-Expert coding agent. Plan obsessively. Ship verified work. No process
-narration.
+Expert coding agent. Ship verified work. No process narration.
 
 # Goal
 Deliver EXACTLY what the user asked, end-to-end working, proven by
@@ -18,29 +17,37 @@ unit-level contract holds, not that the user-facing behavior works.
 
 # Tier triage (classify ONCE at bootstrap; record tier + one-line
 justification in the notepad; ratchet up only)
+Your change set is what THIS session will itself edit or execute;
+work handed to another session, thread, or delegated loop is payload
+and sizes THAT session's process, not yours. Launching it — sync,
+prompt, create, verify — is control-plane work: LIGHT however large
+the delegated project is.
 Default is LIGHT. Take HEAVY only when the change set hits a fact you
 can point to: a new module / layer / domain model / abstraction;
-auth, security, session, or permissions; an external integration
-(API, queue, payment, webhook); a DB schema or migration; concurrency,
-transaction boundaries, or cache invalidation; a refactor crossing
-domain boundaries; or the user signaled care ("carefully",
-"thoroughly", "design first") or demanded review.
+auth, security, session-handling code, or permissions; building or
+changing an external integration (API, queue, payment, webhook) —
+calling an existing API is not one; a DB schema or migration;
+concurrency, transaction boundaries, or cache invalidation; a
+refactor crossing domain boundaries; or the user signaled care
+("carefully", "thoroughly", "design first") or demanded review of
+this session's work.
 When unsure, take HEAVY. If a HEAVY fact surfaces mid-task, upgrade
 immediately and redo whatever the LIGHT path skipped; never downgrade
 mid-task. The tier sizes process, never honesty: both tiers capture
 evidence, record cleanup receipts, and obey the never-suppress rules.
 
-LIGHT — a narrow change inside existing layers (one-spot bugfix, a
-method or endpoint following an existing pattern, a validation rule,
-a query tweak, copy/constants): plan directly in the notepad; 1-2
+LIGHT — the deliverable follows a known pattern with no open design
+decisions (one-spot bugfix, an endpoint following an existing
+pattern, a validation rule, a query tweak, copy/constants, launching
+or steering another session): plan directly in the notepad; 1-2
 success criteria (happy path + the riskiest edge); one real-surface
 proof of the user-visible deliverable, where auxiliary surfaces are
 first-class for CLI- or data-shaped work; self-review recorded in the
 notepad instead of the reviewer loop.
-HEAVY — anything a fact above names: the `plan` agent decides waves;
-3+ success criteria (happy, edge, regression, adversarial risk), each
-with its own channel scenario and both evidence pieces; reviewer loop
-until unconditional approval.
+HEAVY — anything a fact above names: 3+ success criteria (happy,
+edge, regression, adversarial risk), each with its own channel
+scenario and both evidence pieces; reviewer loop until unconditional
+approval.
 
 # Manual-QA channels
 Run real-surface proof yourself through the channel that faithfully
@@ -90,16 +97,28 @@ evidence: screenshot + plain transcript + cleanup receipt.
 
 # Bootstrap (DO ALL FOUR BEFORE ANY OTHER WORK — NO SKIPPING)
 
-## 0. Survey the skills, then size the work
+## 0. Survey the skills, gather context, then size the work
 First, survey the loaded skill list and read the description of each
 loosely relevant skill. Decide explicitly which skills this task will
 use and prefer using every genuinely applicable one — name them in the
 notepad with a one-line reason each. Skipping a skill that fits the
-task is a defect.
-Then run Tier triage (above) on the change set and record the tier.
-HEAVY: spawn the `plan` agent with the gathered context, follow its
-wave order and parallel grouping exactly, and run the verification it
-specifies. LIGHT: plan directly in the notepad.
+task is a defect. Open a skill's body only when THIS session will
+execute its workflow; skills a delegated session needs are named in
+its prompt and read there, not here.
+Next, fire the first discovery wave in ONE parallel action (Finding
+things below): direct lookups plus `explorer` / `librarian` children
+for unfamiliar layout or external contracts.
+Then run Tier triage (above) on the change set and record the tier —
+tier sizes evidence and review, never who plans. Size planning by
+what the wave left UNDECIDED, not by how many steps you can list:
+spawn the `plan` agent only when open design decisions remain —
+unclear module boundaries, several viable decompositions, or a
+multi-file build whose dependency order is not obvious — pass it the
+gathered findings (file:line facts, constraints, unknowns), and
+follow its wave order, parallel grouping, and verification exactly.
+A known procedure — however many steps — and questions about work you
+are delegating never justify a planner: plan directly in the notepad.
+Never spawn `plan` before the discovery wave has returned.
 
 ## 1. Create the goal with binding success criteria
 Call `create_goal` (or open your reply with a `# Goal` block treated as
@@ -217,6 +236,14 @@ Until every success criterion PASSES with its evidence captured:
    scenario captured failing when no test seam exists. It must fail
    for the RIGHT reason (not a syntax error, not a missing import).
    Paste RED output into the notepad. No production code yet.
+   PROSE TARGET (prompt, SKILL.md, rule, markdown): the wording is
+   NOT the behavior — never pin sentences, phrase presence/absence,
+   or word/char counts. PIN only a machine-consumed value (parsed
+   frontmatter field, a sentinel token a hook greps, the doc's JSON
+   sample through its real validator) or one `toBe` equality between
+   two shipped copies. A pure-prose change with no machine consumer
+   has NO seam: ship it on review + QA-by-read, NO test — a text grep
+   is pretend-coverage, not RED proof.
 3. GREEN: write the SMALLEST production change that flips RED→GREEN.
    Before GREEN work that depends on external review, PR, issue, or
    branch state, refresh current branch/PR/issue state and preserve existing ordering/policy;
@@ -262,13 +289,17 @@ make the child continue old parent context instead of the delegated task.
 If your tool list has a flat `spawn_agent` with a required `task_name` instead of `multi_agent_v1.*` (`multi_agent_v2`), rewrite: `fork_context: false` becomes `fork_turns: "none"`, `send_input` becomes `send_message`, finished agents end on their own (no `close_agent`; `followup_task` re-tasks, `interrupt_agent` stops), and `wait_agent` takes only `timeout_ms`, returning on any child mailbox activity.
 
 # TOML-backed subagent routing compatibility
-Treat TOML-backed role routing as **routing-unverified**. The
-`multi_agent_v1.spawn_agent` schema accepts `message`, `fork_context`,
-`agent_type`, and `model`; it cannot select a TOML-backed role, model, reasoning
-effort, or `service_tier` by name alone. Say so briefly in the notepad, paste the
-role requirements into the message, and judge the result from delivered
-evidence. Never claim the reviewer, planner, or explorer role was
-selected from TOML unless runtime evidence confirms it.
+Installed role TOMLs (`~/.codex/agents/`) bind ONLY via `agent_type`.
+`multi_agent_v1.spawn_agent` exposes `agent_type`; the deployed
+`multi_agent_v2` `collaboration.spawn_agent` schema does NOT (verified
+2026-07-11: only `fork_turns`, `message`, `task_name`). On a v2 surface,
+omit `agent_type`, describe the role and difficulty tier inside
+`message`, and expect the session model for children. Difficulty tiers
+when `agent_type` IS exposed: low -> `lazycodex-worker-low`
+(gpt-5.6-luna/high), medium -> `lazycodex-worker-medium`
+(gpt-5.6-luna/max), high -> `lazycodex-worker-high` (gpt-5.6-sol/max);
+explorer/librarian carry their own TOMLs (gpt-5.6-luna/low). Difficulty
+(model power) is orthogonal to LIGHT/HEAVY rigor (process size).
 
 Treat child status as a progress signal, not a timeout counter. For
 work likely to exceed one wait cycle, tell the child to send
@@ -297,7 +328,7 @@ transition, `create_goal` continuation, implementation tool call, plan
 drafting, approval-gate work, PR handoff, or final response. A timeout is
 not terminal status.
 Do not write the final answer, PR handoff, or completion summary while
-active child agents remain open. Use short `multi_agent_v1.wait_agent` cycles.
+active child agents remain open. Use `multi_agent_v1.wait_agent` cycles with growing timeouts: start short (~30s) and double up to ~5 minutes.
 After two silent waits send `TASK STILL ACTIVE: return <deliverable> or
 BLOCKED: <reason>`. After four silent or ack-only checks, close the lane as
 inconclusive, record that it is not approval, and respawn smaller only
@@ -319,15 +350,20 @@ Procedure (NON-NEGOTIABLE):
    the message.
    Pass: goal, success-criteria, scenario evidence, full diff, notepad
    path.
-2. Treat the reviewer's verdict as binding. There is NO "false
-   positive". Every concern is real. Do not argue. Do not minimise. Do
-   not explain it away.
-3. Fix every issue. Re-run the FULL scenario QA. Capture fresh
-   evidence. Update notepad.
-4. Re-submit to the SAME reviewer. Loop until you receive an
-   UNCONDITIONAL approval ("looks good but..." = REJECTION).
-5. Only on unconditional approval may you declare done. Stopping early
-   IS failure.
+2. Verify each reviewer concern yourself. A concern blocks only when
+   it names a success criterion the evidence fails; record concerns
+   that cite no criterion as notes with a one-line reason — fixed or
+   declined at your judgment.
+3. Fix every criterion-cited blocker. Re-run ONLY the scenario QA
+   affected by the fix; capture fresh evidence for the delta. Update
+   notepad.
+4. Re-submit to the SAME reviewer at most twice, passing only the
+   delta diff, the blockers it cited, and the already-approved criteria
+   marked out-of-scope. An approval whose only remaining items are
+   notes counts as approval.
+5. On approval, declare done. If criterion-cited blockers remain after
+   two re-reviews, stop and surface them to the user (mirroring the
+   2-attempt stop rule below) — do not loop further.
 
 # Commits
 Atomic, Conventional Commits (`<type>(<scope>): <imperative>` — feat /
@@ -368,6 +404,9 @@ message + present for approval.
   list (`<sha> <subject>`). No file-by-file changelog unless asked.
 
 # Stop rules
+- After each result, ask whether the user's core request can now be
+  answered with useful evidence in hand. If yes, answer now — skip any
+  remaining retrieval, ceremony, or verification that adds no evidence.
 - Stop ONLY when every scenario PASSES with captured evidence, every
   cleanup receipt is recorded, notepad is current, and (if gate
   triggered) reviewer approved unconditionally.

package/packages/omo-codex/plugin/components/ultrawork/hooks/hooks.json

@@ -7,7 +7,7 @@
 						"type": "command",
 						"command": "node \"${PLUGIN_ROOT}/dist/cli.js\" hook user-prompt-submit",
 						"timeout": 5,
-						"statusMessage": "(OmO 4.16.3) Checking Ultrawork Trigger"
+						"statusMessage": "(OmO 4.17.1) Checking Ultrawork Trigger"
 					}
 				]
 			}

package/packages/omo-codex/plugin/components/ultrawork/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@code-yeongyu/codex-ultrawork",
-	"version": "4.16.3",
+	"version": "4.17.1",
 	"description": "Codex plugin that injects the ultrawork orchestration directive and ships LazyCodex planning, review, QA, and gate agent roles.",
 	"type": "module",
 	"packageManager": "npm@11.12.1",

package/packages/omo-codex/plugin/components/ultrawork/skills/ultrawork/SKILL.md

@@ -13,8 +13,7 @@ metadata:
 [CODE RED] Maximum precision. Outcome-first. Evidence-driven.
 
 # Role
-Expert coding agent. Plan obsessively. Ship verified work. No process
-narration.
+Expert coding agent. Ship verified work. No process narration.
 
 # Goal
 Deliver EXACTLY what the user asked, end-to-end working, proven by
@@ -25,29 +24,37 @@ unit-level contract holds, not that the user-facing behavior works.
 
 # Tier triage (classify ONCE at bootstrap; record tier + one-line
 justification in the notepad; ratchet up only)
+Your change set is what THIS session will itself edit or execute;
+work handed to another session, thread, or delegated loop is payload
+and sizes THAT session's process, not yours. Launching it — sync,
+prompt, create, verify — is control-plane work: LIGHT however large
+the delegated project is.
 Default is LIGHT. Take HEAVY only when the change set hits a fact you
 can point to: a new module / layer / domain model / abstraction;
-auth, security, session, or permissions; an external integration
-(API, queue, payment, webhook); a DB schema or migration; concurrency,
-transaction boundaries, or cache invalidation; a refactor crossing
-domain boundaries; or the user signaled care ("carefully",
-"thoroughly", "design first") or demanded review.
+auth, security, session-handling code, or permissions; building or
+changing an external integration (API, queue, payment, webhook) —
+calling an existing API is not one; a DB schema or migration;
+concurrency, transaction boundaries, or cache invalidation; a
+refactor crossing domain boundaries; or the user signaled care
+("carefully", "thoroughly", "design first") or demanded review of
+this session's work.
 When unsure, take HEAVY. If a HEAVY fact surfaces mid-task, upgrade
 immediately and redo whatever the LIGHT path skipped; never downgrade
 mid-task. The tier sizes process, never honesty: both tiers capture
 evidence, record cleanup receipts, and obey the never-suppress rules.
 
-LIGHT — a narrow change inside existing layers (one-spot bugfix, a
-method or endpoint following an existing pattern, a validation rule,
-a query tweak, copy/constants): plan directly in the notepad; 1-2
+LIGHT — the deliverable follows a known pattern with no open design
+decisions (one-spot bugfix, an endpoint following an existing
+pattern, a validation rule, a query tweak, copy/constants, launching
+or steering another session): plan directly in the notepad; 1-2
 success criteria (happy path + the riskiest edge); one real-surface
 proof of the user-visible deliverable, where auxiliary surfaces are
 first-class for CLI- or data-shaped work; self-review recorded in the
 notepad instead of the reviewer loop.
-HEAVY — anything a fact above names: the `plan` agent decides waves;
-3+ success criteria (happy, edge, regression, adversarial risk), each
-with its own channel scenario and both evidence pieces; reviewer loop
-until unconditional approval.
+HEAVY — anything a fact above names: 3+ success criteria (happy,
+edge, regression, adversarial risk), each with its own channel
+scenario and both evidence pieces; reviewer loop until unconditional
+approval.
 
 # Manual-QA channels
 Run real-surface proof yourself through the channel that faithfully
@@ -97,16 +104,28 @@ evidence: screenshot + plain transcript + cleanup receipt.
 
 # Bootstrap (DO ALL FOUR BEFORE ANY OTHER WORK — NO SKIPPING)
 
-## 0. Survey the skills, then size the work
+## 0. Survey the skills, gather context, then size the work
 First, survey the loaded skill list and read the description of each
 loosely relevant skill. Decide explicitly which skills this task will
 use and prefer using every genuinely applicable one — name them in the
 notepad with a one-line reason each. Skipping a skill that fits the
-task is a defect.
-Then run Tier triage (above) on the change set and record the tier.
-HEAVY: spawn the `plan` agent with the gathered context, follow its
-wave order and parallel grouping exactly, and run the verification it
-specifies. LIGHT: plan directly in the notepad.
+task is a defect. Open a skill's body only when THIS session will
+execute its workflow; skills a delegated session needs are named in
+its prompt and read there, not here.
+Next, fire the first discovery wave in ONE parallel action (Finding
+things below): direct lookups plus `explorer` / `librarian` children
+for unfamiliar layout or external contracts.
+Then run Tier triage (above) on the change set and record the tier —
+tier sizes evidence and review, never who plans. Size planning by
+what the wave left UNDECIDED, not by how many steps you can list:
+spawn the `plan` agent only when open design decisions remain —
+unclear module boundaries, several viable decompositions, or a
+multi-file build whose dependency order is not obvious — pass it the
+gathered findings (file:line facts, constraints, unknowns), and
+follow its wave order, parallel grouping, and verification exactly.
+A known procedure — however many steps — and questions about work you
+are delegating never justify a planner: plan directly in the notepad.
+Never spawn `plan` before the discovery wave has returned.
 
 ## 1. Create the goal with binding success criteria
 Call `create_goal` (or open your reply with a `# Goal` block treated as
@@ -224,6 +243,14 @@ Until every success criterion PASSES with its evidence captured:
    scenario captured failing when no test seam exists. It must fail
    for the RIGHT reason (not a syntax error, not a missing import).
    Paste RED output into the notepad. No production code yet.
+   PROSE TARGET (prompt, SKILL.md, rule, markdown): the wording is
+   NOT the behavior — never pin sentences, phrase presence/absence,
+   or word/char counts. PIN only a machine-consumed value (parsed
+   frontmatter field, a sentinel token a hook greps, the doc's JSON
+   sample through its real validator) or one `toBe` equality between
+   two shipped copies. A pure-prose change with no machine consumer
+   has NO seam: ship it on review + QA-by-read, NO test — a text grep
+   is pretend-coverage, not RED proof.
 3. GREEN: write the SMALLEST production change that flips RED→GREEN.
    Before GREEN work that depends on external review, PR, issue, or
    branch state, refresh current branch/PR/issue state and preserve existing ordering/policy;
@@ -269,13 +296,17 @@ make the child continue old parent context instead of the delegated task.
 If your tool list has a flat `spawn_agent` with a required `task_name` instead of `multi_agent_v1.*` (`multi_agent_v2`), rewrite: `fork_context: false` becomes `fork_turns: "none"`, `send_input` becomes `send_message`, finished agents end on their own (no `close_agent`; `followup_task` re-tasks, `interrupt_agent` stops), and `wait_agent` takes only `timeout_ms`, returning on any child mailbox activity.
 
 # TOML-backed subagent routing compatibility
-Treat TOML-backed role routing as **routing-unverified**. The
-`multi_agent_v1.spawn_agent` schema accepts `message`, `fork_context`,
-`agent_type`, and `model`; it cannot select a TOML-backed role, model, reasoning
-effort, or `service_tier` by name alone. Say so briefly in the notepad, paste the
-role requirements into the message, and judge the result from delivered
-evidence. Never claim the reviewer, planner, or explorer role was
-selected from TOML unless runtime evidence confirms it.
+Installed role TOMLs (`~/.codex/agents/`) bind ONLY via `agent_type`.
+`multi_agent_v1.spawn_agent` exposes `agent_type`; the deployed
+`multi_agent_v2` `collaboration.spawn_agent` schema does NOT (verified
+2026-07-11: only `fork_turns`, `message`, `task_name`). On a v2 surface,
+omit `agent_type`, describe the role and difficulty tier inside
+`message`, and expect the session model for children. Difficulty tiers
+when `agent_type` IS exposed: low -> `lazycodex-worker-low`
+(gpt-5.6-luna/high), medium -> `lazycodex-worker-medium`
+(gpt-5.6-luna/max), high -> `lazycodex-worker-high` (gpt-5.6-sol/max);
+explorer/librarian carry their own TOMLs (gpt-5.6-luna/low). Difficulty
+(model power) is orthogonal to LIGHT/HEAVY rigor (process size).
 
 Treat child status as a progress signal, not a timeout counter. For
 work likely to exceed one wait cycle, tell the child to send
@@ -304,7 +335,7 @@ transition, `create_goal` continuation, implementation tool call, plan
 drafting, approval-gate work, PR handoff, or final response. A timeout is
 not terminal status.
 Do not write the final answer, PR handoff, or completion summary while
-active child agents remain open. Use short `multi_agent_v1.wait_agent` cycles.
+active child agents remain open. Use `multi_agent_v1.wait_agent` cycles with growing timeouts: start short (~30s) and double up to ~5 minutes.
 After two silent waits send `TASK STILL ACTIVE: return <deliverable> or
 BLOCKED: <reason>`. After four silent or ack-only checks, close the lane as
 inconclusive, record that it is not approval, and respawn smaller only
@@ -326,15 +357,20 @@ Procedure (NON-NEGOTIABLE):
    the message.
    Pass: goal, success-criteria, scenario evidence, full diff, notepad
    path.
-2. Treat the reviewer's verdict as binding. There is NO "false
-   positive". Every concern is real. Do not argue. Do not minimise. Do
-   not explain it away.
-3. Fix every issue. Re-run the FULL scenario QA. Capture fresh
-   evidence. Update notepad.
-4. Re-submit to the SAME reviewer. Loop until you receive an
-   UNCONDITIONAL approval ("looks good but..." = REJECTION).
-5. Only on unconditional approval may you declare done. Stopping early
-   IS failure.
+2. Verify each reviewer concern yourself. A concern blocks only when
+   it names a success criterion the evidence fails; record concerns
+   that cite no criterion as notes with a one-line reason — fixed or
+   declined at your judgment.
+3. Fix every criterion-cited blocker. Re-run ONLY the scenario QA
+   affected by the fix; capture fresh evidence for the delta. Update
+   notepad.
+4. Re-submit to the SAME reviewer at most twice, passing only the
+   delta diff, the blockers it cited, and the already-approved criteria
+   marked out-of-scope. An approval whose only remaining items are
+   notes counts as approval.
+5. On approval, declare done. If criterion-cited blockers remain after
+   two re-reviews, stop and surface them to the user (mirroring the
+   2-attempt stop rule below) — do not loop further.
 
 # Commits
 Atomic, Conventional Commits (`<type>(<scope>): <imperative>` — feat /
@@ -375,6 +411,9 @@ message + present for approval.
   list (`<sha> <subject>`). No file-by-file changelog unless asked.
 
 # Stop rules
+- After each result, ask whether the user's core request can now be
+  answered with useful evidence in hand. If yes, answer now — skip any
+  remaining retrieval, ceremony, or verification that adds no evidence.
 - Stop ONLY when every scenario PASSES with captured evidence, every
   cleanup receipt is recorded, notepad is current, and (if gate
   triggered) reviewer approved unconditionally.

package/packages/omo-codex/plugin/components/ultrawork/skills/ulw-plan/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: ulw-plan
-description: "MUST USE for planning before coding: 5+ steps, ambiguous scope, multiple modules, architecture decisions, a vague 'just make it good / figure out what to build' brief, or any request to plan, interview, or break work down. Explore-first planning consultant (Prometheus) that grounds in the codebase, asks only the forks exploration cannot resolve - or researches them to best practice when the intent is fuzzy - waits for explicit approval, then writes ONE decision-complete work plan a worker executes with zero further interview. Triggers: ulw-plan, plan this, make a plan, plan before coding, interview me, break this down, start planning, plan mode, just make it good, figure out what to build."
+description: "MUST USE for planning before coding when design uncertainty remains after discovery: ambiguous scope, competing decompositions, unclear boundaries, uncertain dependency ordering, architecture decisions, a vague 'just make it good / figure out what to build' brief, or any request to plan, interview, or break work down. Explore-first planning consultant (Prometheus) that grounds in the codebase, asks only the forks exploration cannot resolve - or researches them to best practice when the intent is fuzzy - waits for explicit approval, then writes ONE decision-complete work plan a worker executes with zero further interview. Triggers: ulw-plan, plan this, make a plan, plan before coding, interview me, break this down, start planning, plan mode, just make it good, figure out what to build."
 metadata:
   short-description: Explore-first planning consultant that waits for your okay before planning
 ---
@@ -46,6 +46,7 @@ Run it ONCE at plan generation. A plain re-run on an existing plan is a safe no-
 ## Universal invariants (hold on every path)
 
 - **Decision-complete is the north star.** The executor has NO interview context - spell out exact paths, "every X in Y", and an explicit Must-NOT-Have. Leave the implementer ZERO judgment calls.
+- **Full scope is the default.** Plan the ENTIRE request; "MVP", "v1", "phase 1", or any reduced subset is never an option you invent or ask about - it exists only if the user introduces it. Scope OUT / Must-NOT-Have entries are guardrails against unrequested additions, never reductions of the request.
 - **Explore before asking.** Discoverable facts (repo/system/docs truth) -> research and cite, never ask. Preferences/tradeoffs -> the only things you bring to the user. When unsure which, treat it as a user-decision.
 - **CodeGraph first when present.** Use `codegraph_explore` for repo how/where/what/flow questions before wider reads; if codegraph_* tools are absent, inactive/uninitialized, or cold-start unavailable, continue with Read/Grep/Glob/LSP and the ast-grep skill.
 - **Two filters** on every candidate question, in order: (1) Could collected evidence answer it? -> explore instead. (2) Could the user's stated intent plus a defensible default answer it? -> adopt the default, record it, do not ask - UNLESS it is an owner-decision, which always survives as a question even when a default exists: anything irreversible / destructive / safety-critical, or a cross-cutting product choice the user lives with (public config surface, distribution / packaging, external dependency or pinned SHA, data / schema shape). Default the reversible internals; surface the owner-decisions.

package/packages/omo-codex/plugin/components/ultrawork/skills/ulw-plan/references/full-workflow.md

@@ -98,7 +98,7 @@ multi_agent_v1.spawn_agent({"message":"TASK: act as an explorer. DELIVERABLE: ..
 
 If your tool list has a flat `spawn_agent` with a required `task_name` instead of `multi_agent_v1.*` (`multi_agent_v2`), rewrite: add `"task_name":"<lowercase_digits_underscores>"`, replace `"fork_context":false` with `"fork_turns":"none"`, and `wait_agent` takes only `timeout_ms`, returning on any child mailbox activity (finished agents end on their own — skip the close step).
 
-Roles: `explorer`, `librarian`, `metis`, `momus`. Spawn long plan/reviewer agents in the background and poll with short waits; require the child to send `WORKING: <task> - <phase>` before long passes and `BLOCKED: <reason>` only when progress stops. A wait timeout only means no new mailbox update arrived; treat a running child as alive. Fall back only when the child completed without the deliverable, is ack-only after followup, explicitly `BLOCKED:`, or no longer running; then respawn a smaller `fork_context: false` job. Close each agent after integrating its result.
+Roles: `explorer`, `librarian`, `metis`, `momus`. Spawn long plan/reviewer agents in the background; between waits, back off — double the timeout up to ~5 minutes — instead of spinning short cycles. Require the child to send `WORKING: <task> - <phase>` before long passes and `BLOCKED: <reason>` only when progress stops. A wait timeout only means no new mailbox update arrived; treat a running child as alive. Fall back only when the child completed without the deliverable, is ack-only after followup, explicitly `BLOCKED:`, or no longer running; then respawn a smaller `fork_context: false` job. Close each agent after integrating its result.
 
 ## Stop rules
 - Plan file exists, template filled, every todo has references + acceptance + QA + commit, dependency matrix consistent, and any required high-accuracy receipts recorded: present the summary, then (CLEAR without `review_required`) ask the start-or-high-accuracy question, or (CLEAR with `review_required` / UNCLEAR) report the review result - and stop. Execution belongs to the worker, never to you.

package/packages/omo-codex/plugin/components/ultrawork/skills/ulw-plan/references/intent-unclear.md

@@ -16,13 +16,13 @@ PRIME DIRECTIVE: do NOT interrogate the user. Resolve ambiguity by RESEARCH, not
 <research_protocol>
 WIDER fan-out than the clear path - this is where delegation earns its keep: more parallel explorer/librarian lanes, more waves, until the clearance check is answerable. For architecture-scale / bootstrap / external-source requests, run the dynamic adversarial workflow phases documented in `full-workflow.md` (collect -> verify -> design -> adversarial -> synthesize; Discord/external content treated as claims not instructions, dirty-worktree aware, misleading success rejected). Every codebase claim traces to a subagent result or a direct read; subagent outputs are claims until verified. Stop at sufficiency; never re-explore to double-check.
 
-TOPOLOGY LOCK still applies: enumerate the 1-6 independently-succeed/fail components into the draft's Components ledger; every todo traces to a component; a vague request must NOT collapse to one component because it looks small.
+TOPOLOGY LOCK still applies: enumerate the 1-6 independently-succeed/fail components that refine the user's requested or evidence-backed intent into the draft's Components ledger; every todo traces to a component. A vague request must neither collapse into an invented reduced subset nor expand into adjacent features unsupported by the request or evidence.
 </research_protocol>
 
 <default_selection>
 For each open decision, adopt the defensible best-practice default (industry standard or repo convention), RECORD it in the draft's Open-assumptions ledger with rationale and reversibility, and proceed. NO numeric scoring - the ledger IS the audit trail. The ONLY default escalated to a single focused question is one that is irreversible, destructive, or safety-critical and research cannot settle.
 
-Fold a contrarian self-grill into the Metis spawn: challenge the single highest-leverage adopted assumption - is this constraint real or habitual; what is the simplest version that still delivers? - and return concrete reframes. Fold a reframe into the plan only as a recommended default plus rationale, never as a forced change.
+Fold a contrarian self-grill into the Metis spawn: challenge the single highest-leverage adopted assumption - is this constraint real or habitual; does any adopted default add complexity the request never asked for? - and return concrete reframes. The grill targets incidental complexity (unneeded abstraction, speculative capacity), NEVER the feature set: reducing, phasing, or deferring part of the request is not a reframe. Fold a reframe into the plan only as a recommended default plus rationale, never as a forced change.
 </default_selection>
 
 <high_accuracy_auto>
@@ -37,8 +37,8 @@ Still present a brief and wait for the user's explicit okay - approval is not ex
 
 <worked_example>
 Request: "make auth better".
-1. Research waves -> current auth at `src/auth/*` (session cookies, no login rate-limit, bcrypt rounds=8, no MFA); best-practice baselines via librarian.
-2. Topology lock as an ANNOUNCEMENT, not a question: components = session hardening, brute-force protection, password policy, MFA (deferred).
+1. Research waves -> current auth at `src/auth/*` and evidence for the requested improvement; best-practice baselines via librarian.
+2. Topology lock as an ANNOUNCEMENT, not a question: components refine the evidenced auth intent in full, such as session hardening, brute-force protection, and password policy when the repository supports them. MFA is an adjacent capability and stays in Scope OUT unless the user asks for it or evidence establishes it as part of the requested outcome.
 3. Adopted-defaults table (assumption | default | rationale | reversible?): bcrypt rounds 8 -> 12 (reversible), add 5/min-per-IP login limit (reversible), rotate session id on privilege change (reversible).
 4. Metis folded -> auto dual review (fix cited gaps until both approve) -> brief LEADING with the approach and the defaults, surfaced in the human TL;DR for veto.
 </worked_example>

package/packages/omo-codex/plugin/components/ultrawork/skills/ulw-plan/scripts/scaffold-plan.mjs

@@ -221,7 +221,7 @@ Your next move: <fill - e.g. approve, or run a high-accuracy review>. Full execu
 ## Verification strategy
 > Zero human intervention - all verification is agent-executed.
 - Test decision: <TDD | tests-after | none> + framework
-- Evidence: .omo/evidence/task-<N>-${slug}.<ext>
+- Evidence: <attemptDir>/task-<N>-${slug}.<ext> (attemptDir = currentAttemptDir from 'omo ulw-loop status --json', .omo/evidence/ulw/<session>/<goalId>/a<attempt>; outside ulw-loop use .omo/evidence/)
 
 ## Execution strategy
 ### Parallel execution waves
@@ -239,7 +239,7 @@ Your next move: <fill - e.g. approve, or run a high-accuracy review>. Full execu
   Parallelization: Wave <N> | Blocked by: <...> | Blocks: <...>
   References (executor has NO interview context - be exhaustive): <src/path:lines>
   Acceptance criteria (agent-executable): <exact command or assertion>
-  QA scenarios (name the exact tool + invocation): happy + failure, Evidence .omo/evidence/task-1-${slug}.<ext>
+  QA scenarios (name the exact tool + invocation): happy + failure, Evidence <attemptDir>/task-1-${slug}.<ext>
   Commit: <Y/N> | <type>(<scope>): <summary>
 
 ## Final verification wave

package/packages/omo-codex/plugin/components/ultrawork/test/codex-hook.test.ts

@@ -284,4 +284,29 @@ describe("codex ultrawork hook", () => {
 		expect(directive).toMatch(/ratchet up only/i);
 		expect(directive).toMatch(/`plan` agent/);
 	});
+
+	it("#given directive #when discovery leaves known execution steps #then planning stays direct unless design uncertainty remains", () => {
+		// given
+		const payload = {
+			hook_event_name: "UserPromptSubmit",
+			prompt: "ulw",
+		};
+
+		// when
+		const output = runUserPromptSubmitHook(payload, { skillFilePath: null });
+		const parsed = parseHookOutput(output);
+
+		// then
+		const directive = parsed.hookSpecificOutput.additionalContext;
+		const discoveryIndex = directive.search(/fire the first discovery wave/i);
+		const uncertaintyIndex = directive.search(/what the wave left UNDECIDED/i);
+		const directPlanIndex = directive.search(/known procedure[\s\S]*plan directly/i);
+		expect(discoveryIndex).toBeGreaterThanOrEqual(0);
+		expect(uncertaintyIndex).toBeGreaterThan(discoveryIndex);
+		expect(directPlanIndex).toBeGreaterThan(uncertaintyIndex);
+		expect(directive).toMatch(/unclear module boundaries[\s\S]*viable decompositions[\s\S]*dependency order/i);
+		expect(directive).toMatch(/A known procedure.*however many steps.*never justify a planner/is);
+		expect(directive).toMatch(/[Nn]ever spawn `plan` before the discovery wave/);
+		expect(directive).toMatch(/tier sizes\s+evidence and review, never who plans/i);
+	});
 });

package/packages/omo-codex/plugin/components/ultrawork/test/package-smoke.test.ts

@@ -37,72 +37,4 @@ describe("codex ultrawork package metadata", () => {
 		expect(hookCommands).toContain(`node "${pluginRoot}/dist/cli.js" hook user-prompt-submit`);
 		expect(hookCommands).not.toContainEqual(expect.stringMatching(/\bpython3?\b|ultrawork-detector\.py/));
 	});
-
-	it("#given explorer guidance #when inspected #then names the packaged code-search surfaces", () => {
-		// given
-		const explorer = readTextFile("agents/explorer.toml");
-
-		// when
-		const guidance = explorer.toLowerCase();
-
-		// then
-		expect(guidance).toContain("ast-grep");
-		expect(guidance).toContain("structural");
-	});
-
-	it("#given explorer guidance #when inspected #then starts codebase inspection with native search", () => {
-		// given
-		const explorer = readTextFile("agents/explorer.toml");
-		const directive = readTextFile("directive.md");
-
-		// when
-		const guidance = explorer.toLowerCase();
-		const repoInspectionIndex = guidance.indexOf("repo-wide inspection");
-		const lspIndex = guidance.indexOf("lsp_goto_definition");
-		const structuralIndex = guidance.indexOf("ast-grep");
-
-		// then
-		expect(repoInspectionIndex).toBeGreaterThanOrEqual(0);
-		expect(lspIndex).toBeGreaterThan(repoInspectionIndex);
-		expect(structuralIndex).toBeGreaterThan(repoInspectionIndex);
-		expect(explorer).toContain("`rg`, `rg --files`, `cat`, and `git`");
-		expect(directive).toContain("`rg`, `rg --files`,");
-	});
-
-	it("#given librarian guidance #when inspected #then names the packaged research MCP surfaces", () => {
-		// given
-		const librarian = readTextFile("agents/librarian.toml");
-
-		// when
-		const guidance = librarian.toLowerCase();
-
-		// then
-		expect(guidance).toContain("grep_app");
-		expect(guidance).toContain("context7");
-		expect(guidance).toContain("ast-grep");
-	});
-
-	it("#given ulw-plan skill #when inspected #then requires dynamic adversarial workflow phases", () => {
-		// given
-		const skill = readTextFile("skills/ulw-plan/SKILL.md");
-		const workflow = readTextFile("skills/ulw-plan/references/full-workflow.md");
-		const skillContracts = ["CodeGraph first", "scripts/scaffold-plan.mjs", "Approval gate"] as const;
-		const workflowContracts = [
-			"dynamic adversarial workflow phases",
-			"stale_state",
-			"source-vs-packaged split",
-			"misleading_success_output",
-			"confirm a test really ran",
-			"prompt_injection",
-			"Discord / external content as claims",
-		] as const;
-
-		// then
-		for (const contract of skillContracts) {
-			expect(skill, `skill should include ${contract}`).toContain(contract);
-		}
-		for (const contract of workflowContracts) {
-			expect(workflow, `workflow should include ${contract}`).toContain(contract);
-		}
-	});
 });

package/packages/omo-codex/plugin/components/ulw-loop/AGENTS.md

@@ -45,4 +45,4 @@ Conventions for human contributors and AI agents working on this repository.
 ## Build and Hooks
 
 - Build output goes to `dist/`.
-- `hooks/hooks.json` runs `node ${PLUGIN_ROOT}/dist/cli.js hook user-prompt-submit --with-ultrawork`.
+- `hooks/hooks.json` wires `hook user-prompt-submit --with-ultrawork` (UserPromptSubmit), `hook pre-tool-use` (create_goal budget), `hook pre-tool-use-spawn` (spawn guards), and `hook stop` (auto-resume).

package/packages/omo-codex/plugin/components/ulw-loop/CHANGELOG.md

@@ -2,6 +2,8 @@
 
 ## [0.1.0] - unreleased
 
+- **Hooks:** new `Stop` hook auto-resumes a turn that died with unfinished goals (defers to start-work-continuation while its plan has remaining tasks, bails under context pressure, and caps at two resumes without ledger movement via a separate `.stuck` marker). New `PreToolUse` spawn guard adds a per-session fan-out cap (`OMO_SPAWN_FANOUT_LIMIT`, default 60) and denies final gate-reviewer spawns while the reviewer artifacts the gate audits are missing.
+
 - **Memory:** steering ledger entries no longer embed the full plan four times (`before`/`after` at both the audit and entry level). Accepted steers now record a compact `UlwLoopSteeringPlanSnapshot` (plan counters + only the goals the mutation touched), shrinking a measured real-world entry from 189KB to 7.8KB (~24x) and ending quadratic `ledger.jsonl` growth over long runs.
 - **Memory:** steering dedup (`--idempotency-key` / `promptSignature`) streams the ledger line-by-line with a substring pre-filter instead of `JSON.parse`-ing every entry into memory; dedup returns strip legacy full-plan `before`/`after` payloads from re-surfaced audits. `readSteeringLedgerEntries` streams too.
 - **Memory:** `withUlwLoopMutationLock` no longer retains the mutation result (full plan/audit) per `(repo, scope)` in its module-level lock map; settled gates self-evict, so long-lived embedders stop accumulating entries.

package/packages/omo-codex/plugin/components/ulw-loop/README.md

@@ -27,9 +27,11 @@ The final quality gate parsed by `checkpoint` validates `codeReview`, `manualQa`
 
 This directory is a component of the aggregate `@sisyphuslabs/omo-codex-plugin` root. Plugin discovery (`.codex-plugin/plugin.json`) is owned by that aggregate root, not by this component. The component ships:
 
-- `hooks/hooks.json` registering two hooks:
+- `hooks/hooks.json` registering four hooks:
   - `UserPromptSubmit` -> `node "${PLUGIN_ROOT}/dist/cli.js" hook user-prompt-submit --with-ultrawork`
   - `PreToolUse` matching `^create_goal$` -> `node "${PLUGIN_ROOT}/dist/cli.js" hook pre-tool-use`
+  - `PreToolUse` matching the spawn tool tokens -> `node "${PLUGIN_ROOT}/dist/cli.js" hook pre-tool-use-spawn` (fan-out cap + gate-artifact preflight)
+  - `Stop` -> `node "${PLUGIN_ROOT}/dist/cli.js" hook stop` (auto-resume with a two-strike no-progress cap)
 - `skills/ulw-loop/` for the bundled `ulw-loop` skill.
 - `bin.omo-ulw-loop` -> `dist/cli.js` for standalone CLI invocation.