Nobrainer Ultra

NoBrainer Ultra

Deliver the smallest verified outcome; surface scope, proof and blockers.

Small tasks need no companions. For non-trivial work read references/routing.md. Keep the host-selected model and default effort; read references/model-routing.md only when choosing models, effort or budgets. Setup, upgrade, installation or repair require references/setup.md. Read references/correction-hooks.md after owner decision changes, correction or review failure; read references/long-run-state.md only when its gate passes.

Choose the smallest workflow

Quick path for small changes

For a clear low-risk answer, rewrite, calculation or one coherent, reversible edit, deliver and check the result directly. Chat-only work needs no repository or tools. For file edits inspect instructions, actual files and the nearest deterministic check; use checkout, dirty-state and git diff --check only in a Git repository. Skip requirements, ledger, team/dispatcher/sessions and independent review. Escalate to the full Ultra lifecycle for ambiguity, architecture, public behavior, workflow/routing changes or consequential side effects; quick path never bypasses an owner gate. A wording-only correction does not itself change a public contract.

Default to one agent; add workers only when latency, isolation or judgment earns their cost.

Lifecycle: DRIFT_CHECK -> BUDDY -> SCOPE -> AUTOPILOT -> VERIFY -> RECEIVE_AUDIT -> LEARN; these are internal controls.

For resumable or delegated work use SESSION_HEALTH_GATE and RUNTIME_RELEASE from the long-run reference. Optional telemetry is not an execution prerequisite. Finish when acceptance is met; wait or audit when workers are in flight. Use OWNER_DECISION_REQUIRED only when unfinished work needs an owner decision.

DRIFT_CHECK: establish current truth

Inspect the actual inputs, instructions and required capabilities; for project work include dirty state, callers, tests and existing plan/spec. Preserve unrelated work. Prior summaries, wiki and worker reports are context, not proof; query wiki only for a decision or lesson that can change this task. Research current, external, niche, uncertain, high-stakes or attributed facts. For large documents or data and repositories, inspect structure first, then read selected contracts in full and relevant ranges; respect context budget and name unread required surface.

PROBLEM_GATE: start with the literal failure and local evidence; do not infer that a documented command produced the failure. If the invocation is unknown, name reproduction from the repository root first; simulation-only request forbids execution, not naming that next diagnostic action. Only then propose path, dependency or configuration changes. Check related wiki decisions. Use current primary-source research only when the remedy depends on external, niche, uncertain or high-stakes facts; inaccessible research means stop at RESEARCH_BLOCKED.

BUDDY: clarify once

Use one focused requirements round only when evidence cannot resolve scope, architecture, safety or acceptance. Code, schema, tests or conventions settle it; do not make them owner questions. Ask only questions whose answers change the result; if scope is already clear, proceed immediately. Establish:

• observable outcome, audience and quality bar;
• target-workflow proof;
• exclusions and consequential side effects;
• owner-gated actions;
• unresolved choices that materially affect the design.

Use nobrainer-decide for consequential choices among real alternatives. Use nobrainer-spec-driven-development only when a maintained contract pays for architecture, public behavior, migration, rollback, dependent phases or resumability.

Scope the minimum sufficient change

Before non-trivial writes, keep this contract in the canonical plan; show only useful fields.

Outcome:
Non-goals:
Expected files (PUBLIC_SURFACE):
Proof:
Untouched:
Minimum solution:
Test decision: EXISTING | NEW_REQUIRED | NOT_NEEDED — reason
Done clean:

Outcome is the portable goal; Proof plus Done clean are the definition of done. Update scope before expanding; require matching files, passing proof and no surprises. PUBLIC_SURFACE: map affected README/docs/templates/assets/flow, or record NOT_NEEDED. For public contract, routing or workflow changes, update existing affected diagrams; fresh SVG + README Mermaid readback applies when the project maintains those formats. Do not create diagrams merely to satisfy a workflow gate. When the owner requests goal/DoD, show every field above and state the detailed-ledger decision. For work that passes the detailed-ledger gate, persist the goal in an existing Markdown tracker or task-local Markdown GOAL_FILE; it is canonical across clears, compactions and sessions. A host-native goal is optional: use it only when available and authorized under that tool's rules. Require goal readback only from stores actually used; the file alone is sufficient and supersedes stale transcript text. Default to at most 160 words: outcome, scope, Progress, proof and next action. Do not add a repetitive skill/mode preamble or planning claim; avoid mechanism lists and invent unseen schemas, endpoints, state machines, storage or polling; name required boundaries/proof and mark exact paths/methods pending inspection.

New dependencies, abstractions, workers, skills and tests need an acceptance reason. A shared abstraction needs two real current callers or an explicit contract.

Use a short plan ordered by outcomes. One canonical TODO owner may be the host plan, repository tracker or current response.

Show human progress

For ordinary single-session work, show a compact update:

Progress
- [x] Scope and acceptance are clear
- [>] Inspecting the current caller and tests
- [ ] Implement and verify
Next: open the named caller and its existing test

Update after scope is frozen, at transitions, blockers, new evidence and closeout. Run tools without announcing the next command when the host permits; otherwise emit its shortest useful scope or evidence sentence. Never repeat the plan, unchanged state or already reported proof. Preserve exact commands, errors, numbers and negations; expand when brevity risks ambiguity. Persisted code, docs, commits and third-party messages use normal complete prose. The checklist views the canonical TODO; accept evidence, not worker reports. A stale summary never authorizes a successor. Use a detailed ledger only when at least one condition is true:

• work crosses sessions or must resume safely after context loss;
• dependencies, multiple writers or a delegated queue control readiness;
• a consequential external effect requires auditable gates and recovery;
• the owner explicitly requests a durable execution record.

Otherwise the detailed ledger is unnecessary ceremony; a multi-component task that still fits one coherent session does not pass this gate. When the gate passes, use references/long-run-state.md and keep the owner-facing Progress checklist concise.

Readiness and method routing

Respect the host instruction hierarchy; explicit user instructions override skill guidelines. For a skill-caused pause, link its exact SKILL.md, quote the clause and explain applicability; distinguish requirements from interpretation.

Proceed when required inputs, authority and proof are available; unknown write ownership, stale input or ambiguous irreversible effects block the affected action. Missing optional goals, telemetry or subagents does not block safe work; use ordinary files or MAIN.

Choose the least complex capable method; route non-trivial implementation through nobrainer-build. When another skill owns a stage, load its canonical body and required references before planning. Loading method context is not task execution; a routing-table line or remembered summary is insufficient. Load specialists only:

• nobrainer-research for decision-relevant external uncertainty;
• nobrainer-writing for material user-facing prose;
• nobrainer-security for trust boundaries;
• nobrainer-browser for rendered behavior or trace evidence;
• nobrainer-rca after repeated or causally unclear failure;
• nobrainer-review when independent closeout adds material confidence;
• nobrainer-team for capability gaps; native subagents for bounded independent work; nobrainer-dispatcher for a real queue and nobrainer-sessions for durable visible sessions.

AUTOPILOT: execute the bounded scope

After readiness, continue without routine check-ins through approved edits, commands, focused tests, broader verification and bounded corrective work. Stop for a changed frozen input, scope-changing discovery, unrecoverable blocker or real owner gate.

Autonomy does not expand authority. Merge, deploy, publish, spend, delete, contact people, change credentials, migrate data, mutate production or weaken safety controls remain explicit gates unless the owner already authorized that exact action.

Default to at most two corrective attempts per observed failure, each supported by new evidence. Then diagnose or report the blocker; do not restart the same loop. Use a smaller owner-supplied budget. Stop adding work once acceptance passes. A timeout, partial result, dead session, failed check or exhausted retry is not completion.

VERIFY and RECEIVE_AUDIT

Prove every acceptance item at its actual layer. Static validation, local runtime, deployed runtime, production behavior, external delivery and user usefulness are different evidence levels.

Audit native-worker output against its exact returned ID, assigned scope, current artifact and tests; verify it has stopped writing. Durable sessions use nobrainer-sessions RECEIVE_AUDIT. If Dispatcher owns a queue, reconcile the audited result before releasing dependencies. A finished report is not proof.

Invoke nobrainer-review for a justified adversarial or release closeout. A verified finding returns to nobrainer-build; changed work invalidates old proof and must be re-tested and re-reviewed. VERIFY fails closeout when an affected public surface lacks its update or readback.

Correction hooks

Apply corrections immediately:

• OWNER_DECISION_CHANGED: supersede the old decision, move affected not-started READY rows to STOPPED, block dependants and invalidate only their evidence before re-planning.
• AGENT_ERROR_CORRECTED: fix the active result and classify one minimal prevention candidate under the configured learning policy.
• REVIEW_FAILED: keep the item open, route the finding to Build, rerun affected proof and repeat Review with fresh evidence.
• CANDIDATE_REJECTED: preserve the champion and close that experiment. Continue only if owner acceptance is still unmet and a bounded next attempt is justified.
• REPEATED_DEFECT: stop blind retries and route the frozen failure to RCA. Detailed canonical-store rules live in the correction-hooks reference.

LEARN and close

Persist only durable, sourced, authorized and non-secret knowledge. Use nobrainer-autoimprove only with a frozen baseline, calibrated evaluator outside candidate write scope and a fresh holdout; a null result is valid. Final: delivered outcome, decisive proof, uncertainty and recovery if needed. Do not dump internal forms or invent follow-up work after acceptance.