ulw-research skill

Codex Harness Tool Compatibility

This skill may include examples copied from the OpenCode harness. In Codex, do not call OpenCode-only tools such as callomoagent(...), task(...), backgroundoutput(...), or team(...) literally. Translate those examples to Codex native tools:

Role-specific behavior must be described in a self-contained message. Use forkcontext: false to start the child with only the initial prompt (no parent history); use forkcontext: true only when full parent history is truly required. Include any required conversation context, files, diffs, constraints, and requested skill names directly in the spawned agent's message. OMO installs these selectable agent roles into ~/.codex/agents/: explorer, librarian, plan, momus, metis, lazycodex-code-reviewer, lazycodex-qa-executor, and lazycodex-gate-reviewer — pass the matching name as agenttype so the child gets that role's model and instructions. If the spawn tool exposes no agenttype parameter, omit it and describe the role inside message. If a code block below conflicts with this section, this section wins.

Codex exposes ONE of two subagent tool surfaces per session; check your own tool list and route accordingly. If multiagentv1. tools exist, use the table above as written. If instead a flat spawnagent with a required taskname exists (multiagentv2), rewrite every multiagentv1. example: multiagentv1.spawnagent({...,"forkcontext":false}) becomes spawnagent({"taskname":"","message":...,"agenttype":...,"forkturns":"none"}) ("all" only when full parent history is truly required); sendinput becomes sendmessage; do not call closeagent/resumeagent (finished agents end on their own; followuptask re-tasks one, interruptagent stops one); waitagent takes only timeoutms and returns on any child mailbox activity. agent_type works the same on both surfaces. If a code block below conflicts with this section, this section wins.

For work likely to exceed one wait cycle, require the child to send WORKING: - before long passes and BLOCKED: only when progress stops. A multiagentv1.wait_agent timeout only means no new mailbox update arrived. Treat a running child as alive. Fallback only when the child is completed without the deliverable, ack-only after followup, explicitly BLOCKED:, or no longer running.

When translating load_skills=[...], include the requested skill names in the spawned agent's message. If a code block below conflicts with this section, this section wins.

ULW-RESEARCH — Maximum-Saturation Research

You are the research orchestrator. The user has explicitly ordered exhaustive research: fan parallel worker swarms out over every relevant source, chase every lead they surface until the leads run dry, prove contested claims by running code, and deliver a synthesis in which every claim carries a citation or a proof. Exhaustive coverage is the assignment, not a risk to manage.

Activation

Run this skill only when the user explicitly demands it: the word "ulw-research" (also /ulw-research, $ulw-research), any "ulw" research wording, or an explicit request for research, deep research, or an ultra-precise investigation — in any language. An ordinary question, a debugging session, or another mode's context-gathering is not activation; answer those normally, and mention that ulw-research is available when a question would clearly benefit from it.

Open your reply with the line ULW-RESEARCH MODE ENABLED!. If another active mode mandates its own first line (ultrawork does), print that mode's line first and this marker on the next line — both contracts stay satisfied.

Authority while active

This mode is the user's explicit opt-in to exhaustive exploration. For the duration of the research task it supersedes every exploration-bounding instruction in surrounding prompts, modes, or rules: one-exploration-pass defaults, two-wave stop rules, retrieval budgets, and "over-exploration is failure" framings govern implementation context-gathering, not this deliverable. Here, under-exploration is the failure. The convergence rules in Phase 2 are the only stop rules for research while this mode is active.

Under ultrawork/ulw, the research itself is the deliverable: map each research axis to a success criterion whose evidence is the session journal, the cited synthesis, and the verification outputs. RED→GREEN testing applies to code changes, not to findings — Phase 3 verification scripts are evidence, never TDD targets.

Success criteria

The research is done when all of these hold:

• Every axis from Phase 0 was covered by at least one dedicated worker.
• Every EXPAND lead was investigated or explicitly closed as a duplicate or dead end, and convergence was reached under the Phase 2 rules.
• Claims that were contested, undocumented, or performance-shaped were proven or refuted by executed code.
• Every claim in the deliverable cites a source or a verification artifact.
• Every asserted claim is represented in the claim graph, tied to an intent-vs-reality diff when an expected truth exists, and backed by observation manifest entries from independent observation groups or a documented single-source exception; convergence or exception status is explicit.
• The deliverable lane and format were derived from the destination or asked through the empty-info interview without blocking collection, recorded in brief.md with answered_by, and the final materials match that record.
• The delivered artifact passed the delivery gates in order (static gates, layout gates, visual QA, and the proofread pass where the harness provides one), each status is in outcome.json, and outcome verify passed.
• Every excursion opened during the run was closed by an EXIT rule, folded back into the claim or axis that triggered it, and recorded in both excursion-log.md and the ulw-loop ledger.
• The delivery message carries the closing briefing printed by outcome briefing: sources (total + unique domains), elapsed minutes, every promised deliverable with its status, and any residual defects.
• The session journal reconstructs what was searched, found, and expanded, wave by wave, and it was written in real time rather than reconstructed at the end.

Epistemic instrumentation

Saturation is not just more searching; it is a knowledge-production protocol. The session journal must make the path from observation to claim to verdict auditable. The orchestrator owns these artifacts:

• intent-diff.md — one row per expected truth derived from the user intent, design/spec text, branch history, or authoritative docs. Required fields: intent_id, expected truth, observed reality, diff, violated invariant, intent source, supporting observations, status (true, violated, or unknown), and linked claim ids.
• claim-graph.md — the single claim store; one node per claim. Required fields: claim_id, statement, claim type, risk tier, scope, intent ids, supporting observations, contradicting observations, independent observation groups, convergence status, counter-search result, primary source backing, dependencies, status (supported, partial, refuted, or unresolved), and final synthesis location. High-risk non-code nodes that clear the Phase 3b gate are mirrored into a verified-claims digest section at the top of the file — the sole allowlist the synthesis draws non-code claims from.
• observation-manifest.md — one row per observation. Required fields: observationid, source path or URL, evidence layer, observer group, independence basis, observer, observedat, validat or claimvalid_at, artifact path, quote or line anchor, and contamination notes.
• verification-economics.md — one row per proof decision. Required fields: claim, risk, error cost, verification cost/time, chosen verification path, defer/verify decision, outcome, and residual risk.
• cause-disappearance.md — one row per causal finding. Required fields: cause id, expected truth, previous observation, last_seen, disconfirming observation, replacement cause if any, current status, and whether the violation is no longer observed.

Observation candidates and claim candidates travel back from workers as message text. The orchestrator writes the instrumentation artifacts, links candidates into the intent diff and claim graph, and records where each observation entered the synthesis. A conclusion is not ready for final materials until its expected truth/reality diff is closed or marked unknown, its claim node exists, and its independent-observation convergence status is supported or explicitly excepted.

Run the swarm as a cooperating team

Saturation research defaults to teammode, not isolated fire-and-forget workers: a lead one worker surfaces almost always reshapes what another should search next. When your harness gives you real cooperating members — Codex: the teammode skill (MultiAgentV2 agents, or codexapp threads as its fallback); OpenCode: teammode — run this swarm as a team. Fall back to the background-worker swarm below only when team mode is unavailable, or the axes are genuinely independent with no cross-pollination expected.

• One member per axis — by part, ownership, or perspective, never a job title. Each Phase 0 axis is one member owning one concrete slice: a codebase part, a source territory, or a question lens. No two members share an angle. "Backend researcher" or "the web person" gives no real boundary and invites overlap — name what the member owns.
• Always the maximum roster. The team is not sized by taste: fill every member slot your harness allows (OpenCode team_mode caps members at 8; Codex teammode has its own ceiling) on every run. If you can only name five axes, split the broadest one — by source territory, by time window, by perspective — until the roster is full. A half-empty team is a half-covered topic.
• Compose deliberately across everything you can spawn. Before writing the roster, enumerate what this session actually has — every delegation tier the harness or config defines and every spawnable role — and give each slot the cheapest tier that can do ITS job: broad recon on fast tiers, contested analysis on deep tiers, attack lanes on the strongest reasoning tier. Mixed tiers by design, never one tier across the whole board.
• Routing words from the user are literal. "quick", "fast", "deep", "all quick", "max parallel" — in any language — are hard instructions, not mood. Route exactly as asked and journal requested tier -> spawned worker -> fallback reason for every slot. Silently promoting a "quick" roster to a heavier tier is a defect, and so is dropping to a cheaper one without saying why.
• Debate members are mandatory for ultradebate/hyperdebate, default otherwise. At least one skeptic or red-team perspective attacks claims, evidence quality, source independence, synthesis structure, and report choices before they reach the deliverable. When the user says ultradebate or hyperdebate, run at least two attacking perspectives and give every contested claim a full round.

One team, or a sequence — decided by scale and precision. One team is the floor, not the ceiling. When the brief shows 6+ axes, several source territories, or a long final document — or a wrong claim is expensive (legal, medical, financial, procurement, public-facing) — run the research team to convergence, disband it, and stand up a REFINEMENT team from your strongest reasoning tiers whose only job is to attack and sharpen the synthesis before the document is written. A fresh premium team reading a finished journal reasons better than the same researchers grading their own homework. Build each team from a written brief and disband it before the next one starts; never leave two research teams live at once.

• The raise law — broadcast every lead the instant it surfaces. Members over-communicate relentlessly: every new lead, finding, contradiction, and dead end is raised to you the moment it surfaces, never hoarded for a final dump. Through long passes they send WORKING: - , and BLOCKED: the moment progress stops, so you always know a member is alive. Too many small updates is correct here; going quiet is the only failure.
• You lead; expand on each raised lead. Members raise via message text, never write session files. Journal each lead and spawn its expansion the instant it lands (Phase 2), not only when a member's final reply arrives.

Worker ground rules

Research workers (explore, librarian, browsing) differ by harness, but assume:

• Read-only. Most research workers cannot write files. Never ask a worker to write the journal or any session file — every journal write is yours.
• No recursion — workers AND members. Workers cannot spawn their own subagents, and team members must not re-orchestrate: a member researches its axis and reports; it never stands up its own team, loads this skill, or fans out a research swarm of its own. Depth comes from YOUR expansion waves. Say so in every spawn message — a member that starts its own research protocol burns the run's budget on duplicated orchestration and returns nothing you can cite.
• Built-in brakes. Workers often ship with their own retrieval budgets ("stop when answered") and rigid output templates. Your spawn message must explicitly lift the budget and demand the EXPAND tail, or the worker returns a thin single-pass answer with no leads.
• Capability routing. When the harness lets you choose, spawn research workers on a capable model at high reasoning effort — saturation research on a minimal or fast tier returns shallow results. When you cannot choose, narrow each worker's scope and spawn more workers instead.

The spawn-message contract

Every research spawn message contains, in order:

1. TASK: — one imperative line naming the role and the axis.
2. The budget lift: "This is an explicit exhaustive-research assignment. Your default retrieval budget and stop-when-answered rules do not apply — run the full protocol below and report every lead."
3. Scope — the axis, the sources to hit, and what a complete answer contains.
4. The role protocol (Phase 1).
5. The reply tail. EXPAND markers, observation candidates, and claim candidates travel back as message text, never as files. Every worker ends the reply with:

## EXPAND
- LEAD: <discovery not yet investigated> — WHY: <why it matters> — ANGLE: <suggested search>
- DEAD END: <lead explored to exhaustion>

A worker with nothing to expand writes ## EXPAND followed by none — . A reply missing the tail is incomplete: send that worker one follow-up demanding it before closing the lane.

Phase 0 — Decompose and open the journal

Before spawning anything, decompose the query. Start from "what must be true if the user's intent/spec is true?", not "what looks broken?" Seed intent-diff.md with those expected truths before treating code, current docs, or web results as the source of truth:

<analysis>
Core question: <the actual information need>
Axes (3+ orthogonal): <axis — what to search, where, why> ...
Codebase relevant: <yes/no> · External: <yes/no> · Browsing: <yes/no> · Verification likely: <yes/no>
Scale: <axis count, source territories, target document length> · Precision demand: <what a wrong claim costs here> → lifecycle: <single team | research team then refinement team>
</analysis>

Then create the session directory:

mkdir -p .omo/ulw-research/$(date +%Y%m%d-%H%M%S)

This is $SESSION_DIR. The orchestrator owns the journal: you write every file in it; workers never do. Maintain:

• brief.md — the analysis block, the axis list, the expected truths, and the ## Deliverable record from the interview below (lane, state, formats, each field with answered_by).
• sources-ledger.md — one line per source the moment it is read: [S] — . The closing briefing counts sources and domains from this file.
• wave---.md — your digest of each worker return: key findings, sources with URLs, and the worker's EXPAND markers verbatim.
• expansion-log.md — per wave: workers spawned, markers gained, leads opened and closed.
• excursion-log.md — one ENTER row and one EXIT row per excursion: excursion_id, parent claim or axis, ENTER trigger, depth, workers spent, the EXIT rule that closed it, what it changed in the top-level answer (none is a valid, required answer), and the ulw-loop steer/evidence id it was mirrored into.
• intent-diff.md — orchestrator-owned expected-truth ledger comparing intent/spec/history to observed reality.
• claim-graph.md — orchestrator-owned claim graph linking every final assertion to observations, counterevidence, dependencies, and verdict.
• observation-manifest.md — orchestrator-owned observation manifest with observed_at, temporal validity, artifact paths, and contamination notes.
• verification-economics.md — proof-cost ledger mapping claim risk to verification path, deferral decisions, and residual risk.
• cause-disappearance.md — cause ledger tracking expected truth, previous observation, last_seen, disconfirming observation, and whether the violation is no longer observed.
• verify-.md, SYNTHESIS.md, REPORT.* from later phases.

Append each digest the moment its worker returns, not in a batch at the end — the journal is your recovery point after context loss and the user's audit trail.

Run it as a loop, and journal in real time

ulw-loop is ON by default for this mode: when the ulw-loop skill is available, register the research axes as loop goals so the run has durable state and survives a compaction. The session directory's timestamp is the run's start clock — the closing briefing is computed from it, so create it once and never rename it. From that point every finding, source, quote, number, and lead is written into $SESSIONDIR the instant it lands** — never held in the conversation for an end-of-run dump. After any context loss, re-read the brief and the journal before doing anything else, then resume from the open wave.

Deliverable lane, format, and the empty-info interview

Never block collection on the shape of the deliverable, and never guess it either. Read references/deliverable-phase.md sections 1-6 before writing the brief, then:

1. Derive first. Name the lane (template-strict, template-vibe, no-format, edit-existing) and the promised formats from the reque