forge-mcp
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@forge-mcpstart a pipeline to add a login page to my app"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
forge-mcp
The director of the forge pipeline. An MCP server that tells Claude Code which phase to run next and what it must produce, then validates the evidence before letting the flow advance. Claude is the executor β forge never runs the work itself. It directs, you build.
πͺπΈ LΓ©elo en espaΓ±ol
The idea
A long piece of work β design a feature, build it, test it, ship it β is easy to do out of order, skip a step of, or declare "done" without proof. forge turns that flow into a pipeline of phases that can only advance in order, and only when each phase hands back real evidence.
forge says which phase is next and exactly what it expects from it (the full criteria, not a summary).
You (Claude) execute the phase β reading code, writing it, running tests, asking the user when a decision is theirs.
You report the phase done with evidence; forge validates it and advances.
You cannot skip a phase, cannot close one with empty or fake evidence, and cannot finish before every phase is closed.
The state lives in SQLite (node:sqlite, a Node built-in β no native build step), so a
run resumes in any session: Claude's context does not survive a close, a compaction or
picking up the next day β the flow's state does.
Related MCP server: Vibe-Coder MCP Server
The 13 phases
classify β clarify β setup β precondition β design β plan β build
β gates β qa β reconcile β contraste β reflect β deliverPhase | What it does | Asks the user | Optional |
| Classify the request's nature (QUESTION / MICRO / STANDARD / HIGH-RISK) and scope. | ||
| Detect ambiguities that change the product; bring the important ones to the user with options and consequences. | yes | |
| Decide the stack and its REAL versions (via | ||
| Verify the conditions to safely start the build are actually met (tools present, env ready). | ||
| Before coding: brainstorm the solution, a UX/UI design brief, a QA plan with edge cases, a code-quality guide β persisted as artifacts. | ||
| Decompose the work into atomic tasks with disjoint file ownership, grouped into independent blocks. | ||
| Implement the plan. You build β reusing what exists, never rewriting a whole file, respecting the strict linter. | ||
| Run the repo's real gates (strict lint + build + tests). Truth is the exit code, not the model's self-report. | ||
| Verify for real: run the app end to end, then ATTACK it (odd inputs, limits, impossible states) and report what broke. | ||
| Only if parallel work may have duplicated logic or created conflicts β resolve them. | yes | |
| An independent review that does NOT know the QA verdict, exploring the finished build with fresh eyes. | ||
| Look back on how this run went (what passed, what fell back, what failed) and extract lessons. | ||
| Publish per the request (remote push, deploy). Never reports "online" without a real URL; skips explicitly with a reason if it does not apply. | yes |
Each phase carries a goal (short instruction), a full systemPrompt (the complete criteria
for that phase), the skills to load before running it, and flags for whether it needs the
user or is optional. All defined in src/phases.ts.
Evidence is validated, not trusted
forge does not accept "done" as a string. forge_complete_phase validates the evidence
each phase must hand back, and rejects the close if it does not hold up:
gatesrequires the real exit codes (lint / build / test) and they must all be0.qarequires a structured result: it passed, and it was actually attacked.clarify(a user decision) requires an explicituserConfirmed.An optional phase can only be skipped with a stated reason.
So a phase cannot be closed with an invented summary. The pipeline advances on proof.
The skills library (128)
src/skills.ts + the skills/ folder ship 128 skills, each with its own SKILL.md,
versioned in the repo (forge is self-contained β it does not depend on anything external for
these). Each phase declares which skills it loads; the domain map (SKILL_MAP) says which
skills belong to which domain. Skills load on demand β Claude asks for the list with
forge_skills and the content of a specific one with forge_skill(name), never all at once.
The 7 tools
Tool | What it does |
| Start a new run; returns the first phase ( |
| Which phase a run is in and its progress ( |
| The CURRENT phase with its detailed goal, full |
| Close the current phase with a summary and validated evidence, then advance. If it was the last phase, mark the run |
| List active runs β to resume from any session without re-reading context. |
| List the full skills library, or filtered by domain if a phase is given. |
| Return a specific skill's |
Install
git clone https://github.com/DevRik99/forge-mcp
cd forge-mcp
npm install
npm run buildRegister it in Claude Code (.mcp.json or project config):
{ "mcpServers": { "forge": { "command": "node", "args": ["dist/server.js"] } } }Requires Node β₯ 22.5 (for node:sqlite).
How a run resumes
The state lives in a single global SQLite DB at ~/.forge/forge-mcp.db (override with
FORGE_MCP_DB), so every project shares one store and runs are told apart by their cwd.
Two tables:
runs: one row per run (id,request,cwd,current_phase,status, timestamps).phase_artifacts: one row per closed phase (run_id,phase,summary,closed_at) β the real decision/artifact Claude reported, not just a boolean flag.
After a lost session (close, compaction, next day), any new Claude session with this MCP connected can:
Call
forge_tasks()to see which runs are still active and in what phase.Call
forge_next(runId)to get the current phase's fullsystemPromptagain β Claude does not need to remember anything; forge hands it back verbatim.Read the closed phases' artifacts via
forge_statusso nothing already decided (e.g. inclarify) is re-asked.
Guarantees (and honest limits)
forge enforces: the phase order, closing every phase before finishing, and validated
evidence per phase (no fake gates/qa, no skipping user decisions or optional phases
without a reason). Under concurrency, closing a phase is atomic β a stale double-close is
rejected, not silently applied.
forge cannot stop you from editing the project without using it at all β an MCP only
sees its own tools, not your Edit/Write/Bash. To force every change through the
pipeline, pair it with the forge-flow gate from
claude-gates, which blocks edits when no forge
run is active.
License
MIT.
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables comprehensive software lifecycle management with structured tracking of requirements, tasks, and architecture decisions through an SQLite database with full traceability and automated state validation.38MIT
- FlicenseNot gradedqualityNot gradedmaintenanceImplements a structured development workflow for LLM-based coding with feature clarification, PRD generation, phased development, and task tracking. Guides LLMs through organized feature development from requirements gathering to completion with document storage and progress monitoring.30
- FlicenseNot gradedqualityCmaintenanceTransforms ideas into detailed, executable development plans with built-in verification, lessons learned tracking, and GitHub issue remediation workflows. Guides Claude through structured interviews, plan generation, execution with Haiku agents, and verification with Sonnet agents to maintain context and code quality across sessions.6
- AlicenseNot gradedqualityFmaintenanceAn autonomous software-engineering pipeline for Claude Code that runs on real evidence: codebase intelligence tools, git analytics, deterministic verification rules, and zero LLM-judges-LLM. Every claim traces to a paper and every PR passes its own gates.1MIT
Related MCP Connectors
Adaptive plan/build/review cycles for AI coding assistants, persisted across sessions.
Persistent context for Claude. Your AI always knows your projects and next actions across sessions.
Cross-agent artifact workspace with provenance across Claude Code, Codex, Cursor, LangGraph.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/DevRik99/forge-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server