MISSION BRIEFING

Execute the full 13-phase build protocol from PRD to production.

/assembleFuryAll

The full pipeline. Architect → Build → Triple Review → UX → Double Security → DevOps → QA → Test → Crossfire → Council. The review-heavy fan-out phases now run as a dynamic Workflow (ADR-067) over the mission's working diff, keeping the 15+-agent fan-out out of the lead's context; the build phases stay prose orchestration. One command to rule them all.

/campaignSiskoAll

The war room. Sisko reads the PRD, identifies every remaining mission, and executes them one by one — running /assemble for each — until the entire product is complete. Autonomous execution is the default — no flag needed.

/imagineCelebrimborAll

The forge artist. Celebrimbor scans the PRD for visual asset requirements, derives a style from the brand section, and generates images via DALL-E 3. Portraits, illustrations, OG images, hero art — whatever the PRD describes and code can't produce.

/prdSiskoAll

The PRD generator. Sisko runs a 5-act structured interview — product vision, tech stack, features, UI, shipping plan — and produces a complete PRD with valid YAML frontmatter.

/blueprintPicardAll

The Blueprint Path — validate a pre-written PRD, discover supporting docs, merge directives, provision infrastructure, and hand off to /campaign for autonomous build.

/growKelsierFull

Kelsier's 6-phase growth protocol. Audit your product, optimize SEO, create content, run ads, manage outreach, measure results. The forge that grows what it builds.

/treasuryDocksonFull

Dockson's per-project financial operations. Connect revenue sources, allocate budgets, execute spend, reconcile daily. Treasury paths are project-scoped — each project has its own spend-log, revenue-log, and budget files.

/portfolioDocksonFull

Dockson's cross-project financials. Track all VoidForge projects from one place — status, revenue, traffic, ad spend, ROAS.

/cultivationKelsierFull

Plant the growth engine. Cultivation wires up ad platforms, analytics, content pipelines, and social scheduling — the infrastructure Kelsier's crew needs before they can burn.

/currentTuvokFull

Tuvok's Deep Current — autonomous campaign intelligence. Scans the site, analyzes the market, proposes growth campaigns, learns from results. The forge that thinks about what to build next.

/qaBatmanAll

The double-pass. Batman's team probes every path, every boundary, every assumption. Then they do it again. If it breaks, they already know.

/testBatmanAll

Batman writes the tests everyone else forgot. Coverage gaps identified, test architecture designed, missing paths covered. The Dark Knight doesn't leave witnesses.

/sentinelKenobiAll

OWASP audit with parallel and sequential phases and red-team verification. Critical and High findings pass through a vote-based REFUTE Gate — cross-universe skeptics default to REFUTED and must confirm each finding in the actual code before a fix is spent, then re-rate its severity from the votes. Canonical name for the security command (ADR-050).

/securityKenobiAll

OWASP audit with parallel and sequential phases and red-team verification.

/uxGaladrielAll

Adversarial UX/UI review with a11y audit and verification pass.

/engagePicardAll

Picard reads every line against the pattern library. Every Must Fix and Should Fix finding now passes through a vote-based REFUTE Gate — cross-universe skeptics default to REFUTED and must confirm it in the actual code before it reaches the fix batch. A --pre-deploy --diff mode scopes the review to the working-tree diff, auto-sizes the lens panel to the change, and makes the verify pass mandatory. Canonical name for the review command (ADR-050) — coexists with Claude Code's native /review skill.

/reviewPicardAll

Picard reads every line against the pattern library. Compliance, quality, maintainability. If it doesn't meet the standard, it doesn't ship.

/gauntletThanosAll

The ultimate test. 5 rounds, 30+ agents across 9 universes, escalating from discovery to adversarial warfare. Re-platformed onto a dynamic Workflow (ADR-067): discovery → JS dedupe → 3-lens adversarial REFUTE verify → crossfire → council, with the 60–80 agents' findings kept out of the lead's context. Review-only — no build. If your project survives the snap, it's ready for anything.

/assessPicardAll

Picard's pre-build assessment — evaluate an existing codebase before a rebuild, migration, or VoidForge onboarding. Chains architecture review, assessment-mode Gauntlet, and PRD gap analysis into a unified 'State of the Codebase' report.

/aiHari SeldonAll

Seldon's AI Intelligence Audit — reviews every LLM-powered component in your application. Model selection, prompt architecture, tool-use schemas, orchestration patterns, evaluation strategy, safety, token economics, and observability.

/audit-docsCoulsonAll

A lean, code-free audit of the documentation corpus only. A Surfer-led doc roster — Troi, Wong, Irulan, Coulson — hunts four classes of defect: doc-currency drift, broken cross-references, command↔method desync, and version-SSOT inconsistency. Report-only — it reads no application source, runs no build, and writes no fixes.

/devopsKusanagiAll

Kusanagi provisions, deploys, and monitors. Six deploy targets, automatic SSL, health checks that actually check health. The Major doesn't do half-measures.

/architectPicardAll

Picard reviews the big decisions. Schema design, scaling strategy, ADRs. When agents disagree, he resolves the conflict. Make it so — but make it maintainable.

/gitCoulsonAll

Version bump, changelog, commit — full release management. Now with release discipline: the version tag is applied by default (a tagless release commit is invisible to release tooling), and npm publishing is gated behind an explicit --npm opt-in that only ships packages whose version matches the bump, in dependency order.

/voidBombadilAll

Old Tom keeps the forge in tune. Runs voidforge update to pull the latest methodology from npm, and makes sure the tools that build your tools stay sharp.

/thumperChaniFull

Telegram bridge with Gom Jabbar authentication and sandworm relay.

/debriefBashirAll

The field medic. Bashir examines what happened during a build or campaign, diagnoses what went wrong, traces root causes, and writes a post-mortem that prevents the same failure from recurring. Can file upstream issues to improve VoidForge itself.

/dangerroomFuryFORGE LABSFull

Per-project mission control — a tab inside the project dashboard with build progress, agent activity, findings, deploy status, and campaign state. WebSocket broadcasts are project-scoped via subscription rooms. Access controlled by resolveProject() middleware.

/deployKusanagiAll

Kusanagi's deploy agent — target detection, pre-deploy checks, deploy execution, health verification, rollback on failure. Supports VPS, Vercel, Railway, Docker, Cloudflare, and static targets.

/vaultHari SeldonAll

Seldon's Time Vault — distill session intelligence into a portable briefing for session handoff. Preserves decisions, failed approaches, cross-module relationships, and execution plans so the next session starts informed.