An opinionated, disciplined, quality-first software production system. A conveyor that carries value from IDEA to PRODUCTION across role-tuned agents and value-stations (IDEA ▶ ROADMAP ▶ PLAN ▶ EARS ▶ FEATURE ▶ TEST ▶ IMPLEMENT ▶ STORY ▶ SHIP), governed by three pillars — knowledge parity, quality-first with a performance-delta gate, and waste elimination (muda·mura·muri) — under an overarching token-efficiency constraint. Emits markdown; uses the sentinel and pressroom companions automatically when present.
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Verify FOUNDRY's external tool dependencies are installed and reachable — a ✓/✗ table by tier (advisory; --strict to fail on a missing required tool).
Pin unpinned behaviour — find the behaviour least covered by tests, add the missing coordinate(s), and loop until every behaviour is pinned (100% coverage is the floor that results).
Run the FOUNDRY production cycle — ingest the ROADMAP, plan, tier by token budget, and drive items idea→product through the value-flow conveyor.
Inspect the FOUNDRY system itself — audit skills, agents, and knowledge for drift, gaps, and duplication; produce a severity-ranked report.
Detect the current development phase of each IN_PROGRESS roadmap feature and install the right skill for the next stage.
FOUNDRY LEAD ENGINEER — the orchestration architect of a FOUNDRY cycle. Ingests a complete ROADMAP, analyses the full codebase, determines the required stack, identifies shared infrastructure across roadmap items, decomposes every item into parallelisable tasks, estimates token budgets, and produces the FOUNDRY_PLAN.md that drives the entire production cycle. Also creates and maintains the SUBJECT_MATTER_UNDERSTANDING document. Spawned once per FOUNDRY cycle at the start of §5.
Automated behaviour-pinning agent. Finds the behaviour least pinned by tests (the file with the most unpinned lines), adds the missing coordinate(s) that locate its correct implementation, and loops until every behaviour is pinned or the user stops. 100% coverage is the floor that results, never the target — the variable is coverage density (happy/unhappy/abuse). Writes IN_PROGRESS.md for disaster recovery.
Creates and maintains the implementation plan artifact with checklist and resumption instructions. Spawned by the lifecycle-orchestrator or FOUNDRY at the start of a new roadmap item's lifecycle.
Produces EARS specification updates with unique requirement IDs and traceability hooks. Spawned after step-0-plan completes and PLAN_COMPLETE sentinel is present.
Writes .feature behavior documentation with happy, unhappy, and abuse scenarios mapped to EARS IDs. Spawned after EARS_COMPLETE sentinel is present and EARS-REVIEWER has passed.
FOUNDRY — Technical Tooling for Software Evolution. The production orchestrator for this production system. Ingests a complete ROADMAP.md, tiers all items by PRIORITY_STATUS against a token budget, decomposes the full backlog into parallelised expert subagents (PHASE_POOL + VALUE_HANDLER_POOL), drives every item through the full DEV_SYSTEM pipeline (EARS → FEATURE → TEST → IMPLEMENT → STORY), and records every completion in IDEA_COST.jsonl. FOUNDRY is the engine of the idea-to-production conveyor. Trigger when the user says: "run the foundry", "process the roadmap", "orchestrate the roadmap", "ship the backlog", "build everything", "start FOUNDRY", "run a foundry cycle", "what would this cost?", "estimate the backlog", or "inspect FOUNDRY".
Verify that FOUNDRY's external tool dependencies are installed and reachable on this machine — language toolchains, test runners, the Playwright MCP + a browser, debuggers, and language servers. Trigger with /foundry:check (or "check foundry prerequisites", "are my foundry tools installed?"). Runs a fast ✓/✗ probe grouped by tier (required / recommended / optional). Advisory by default (never blocks); pass --strict to fail when a required tool is missing. Reads the canonical manifest skills/check/requirements.tsv.
Deep code quality analysis across Clean Code, SOLID, DRY, Clean/Hexagonal Architecture, TDD, BDD, DDD, 12-Factor App, and Pragmatic Programming principles. Use proactively whenever the user asks about code quality, technical debt, refactoring, architecture review, code review, maintainability, test coverage, coverage gaps, coverage loops, or says things like "is this clean?", "how good is this code?", "review my architecture", "find smells", "improve my tests", or "chase coverage". Also triggers the /coverage-loop slash command, which finds behaviour not yet pinned by a test and adds the missing coordinates (100% coverage is the floor that results, not the target).
Use when orchestrating lifecycle execution of an IDEATOR item across SDLC steps 0-9, including stage maturity checks, quality gates, and iterative progress to Definition Of Done.
The FOUNDER method — how a software production house turns an idea into a product one vertical slice at a time through value-stations and value-handlers, governed by a five-level performance-instrumented test contract. Use when planning the production line, defining stages and their owners, onboarding a new value-handler agent, or explaining WHAT is being built and HOW. Trigger phrases: "founder method", "value stations", "who owns this stage", "production line", "how does our process work".
Modifies files
Hook triggers on file write and edit operations
Uses power tools
Uses Bash, Write, or Edit tools
Carry software from the spark of an IDEA to PRODUCTION — nine composable Claude Code plugins, one disciplined, test-first value flow: discover ▸ refine ▸ design ▸ build ▸ assure ▸ secure ▸ publish ▸ operate.
Start here →
/i2p-helpbrowses every power you have ·/i2p-flowshows the pipeline ·/i2p-reviewgives one verdict from every reviewer. The i2p plugin is the front door; concierge greets whoever opens the repo.
A disciplined, test-first conveyor that carries VALUE from IDEA to PRODUCTION, governed by three pillars under one overarching constraint:
| Pillar | also called | in one line |
|---|---|---|
| 🧭 Knowledge-parity | knowledge-alignment | understand the ask completely before acting — recurring questions become written answers, asked once. |
| 🛡️ Quality-first | quality-confidence | quality is built in, not inspected in — every station, strengthened by a performance-delta gate; a gate is never weakened to make progress. |
| ♻️ Waste-elimination | muda · mura · muri | remove waste in every form, including rediscovery — a bug caught in development is far cheaper than one in production. The three Ms: muda (waste), mura (unevenness), muri (overburden). |
Overarching constraint — token-efficiency: thin skills, fat references; define once, reference many; load only what a station needs. And the marketplace is self-improving: when an element learns from a mistake, it folds the fix back into itself — self-cleaving into smaller, sharper parts where needed — and raises a PR so every user inherits the improvement.
Not decoration — these are the ideas the whole system obeys. The philosophical spine is first-principles.md; the operation is VALUE_FLOW.md.
Knowledge-parity · quality-first · waste-elimination — under one token-efficiency constraint.
A failing test written before the code is a location in logical space; one written after is just a description.
Extract the decidable core; dependencies flow one way, inward. pure ⇒ testable ⇒ coordinate-able ⇒ maximally parallel.
Every pass at least halves the distance to perfection; fix upstream once, ship to all.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimnpx claudepluginhub agentic-underground/idea-to-productionThe FRONT DOOR of the idea-to-production marketplace — the marketplace-level front door the seven specialist plugins never had. Adds /i2p-help (browse the whole suite's powers, grouped by the DISCOVER ▸ IDEATE ▸ DESIGN ▸ BUILD ▸ ASSURE ▸ SECURE ▸ PUBLISH ▸ OPERATE ↻ value flow), /i2p-review (a cross-plugin adversarial review that fans out EVERY installed reviewer — code, design, docs, security — into one verdict), /i2p-check (consolidated readiness across all installed plugins), and /i2p-flow (where each plugin sits in the flow and the next command to run). Introduces itself on session start and surfaces occasional ≤25-word tips. A thin orchestrator: it delegates to the specialist plugins by capability and never re-implements them. Carries the KAIZEN self-improvement covenant.
The DISCOVERY front door of the idea-to-production marketplace. MARKET-SCANNER helps a user find a worth-building opportunity before any code exists: set a standing discovery /discovery-goal, then run /market-scan — an adversarially-challenged dialogue over a market parameter taxonomy (problem severity, demand evidence, market size, willingness-to-pay, pricing power, competition, reachability, and stack-fit) that proposes, scores, validates, and NARROWS candidate opportunities, killing weak ideas early. Loop it over your goal until one candidate earns a keep verdict — the spark. Hands a validated OPPORTUNITY to the ideator plugin (REFINEMENT) when installed, or emits a structured markdown opportunity brief standalone. Self-improving: downstream ideation/build feedback sharpens the parameters and scoring over time.
The REFINEMENT phase of the idea-to-production marketplace. IDEATOR takes a validated opportunity (from the market-scanner plugin) or a raw idea you already have and refines it to knowledge-parity through an adversarially-challenged dialogue, producing the IDEA PACKAGE: precise, high-clarity AGENT-FACING handoff documents (an idea brief, a subject-matter-understanding seed, the first vertical slice, and a handoff contract that satisfies FOUNDRY's discovery exit criteria) plus rich, illustrated USER-FACING documents (the opportunity narrative, a parameter scorecard, market/pricing/competition charts, a user-flow, and a mockup screen) rendered by capability: charts via pressroom, and user-flows/mockups via atelier /mockup (designed to the canon and design-reviewed before you see them) when installed; degrades to markdown otherwise. The package is iterated with the user before hand-off, then carried to PRODUCTION by FOUNDRY. Self-improving: ambiguities a downstream builder hits flow back to sharpen the question-bank and the package contract. Supersedes FOUNDRY's inline ideator (which remains as a graceful-degradation fallback).
The pre-release security gate for any repository. SENTINEL catches PII, secrets/credentials, and supply-chain risk before they ship. It runs parallel, multi-agent audits across data files, git history, source code, and frontend, then consolidates findings into a single severity-ranked report with an overall PASS / REVIEW / BLOCK verdict. Runs standalone on any project, or serves as the SECURITY gate for the foundry plugin when both are installed. Carries a self-improvement covenant: every missed pattern becomes a rule, every false positive becomes an exclusion.
The DESIGN studio — a designer's workshop that both makes and critiques visual work to a commercial-grade standard. /ui-review crawls the navigable routes of any SPA (screenshot + accessibility snapshot per route) and writes a scored, prioritised critique where every finding cites a named principle; /mockup composes polished screens, wireframes, and user-flows and runs them through a convergent designer↔reviewer loop until they clear a design-fitness rubric. The reviewer is a heavyweight expert in the design canon — Gestalt, the UX laws (Fitts/Hick/Miller/Jakob), Nielsen's heuristics, Norman's emotional design, and WCAG 2.2 accessibility. Composes with foundry's frontend design-system and pressroom's graphics by capability; degrades to markdown when absent. Self-improving: design feedback sharpens the canon and the rubric via a PR.
Complete creative writing suite with 10 specialized agents covering the full writing process: research gathering, character development, story architecture, world-building, dialogue coaching, editing/review, outlining, content strategy, believability auditing, and prose style/voice analysis. Includes genre-specific guides, templates, and quality checklists.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.