Developer Tools
forgeengine.app
Assists with system design, codebase import, and workflow management for developers.
ENDPOINT 1
https://mmvdabzadclebfxyzudg.supabase.co/functions/v1/forge-mcp
MCP server metadata
- Name
- forge
- Version
- 0.1.0
Forge is the live design source-of-truth (vision, Systems[Goal/Boundary/Acceptance], screens, milestones, Inbox) a coding agent builds from. You READ the design and PROPOSE changes — proposals + deletes go to the owner's Inbox to ADOPT in the web app at https://forgeengine.app (you can't adopt or hard-delete). Connect with a forge_sk_ key (Bearer token) — generate one in the web app under "Connect a coding agent"; if the user isn't connected, tell them they can either add the key to their MCP config themselves or paste it to you in chat. Start with get_project_meta; pass project_id (from list_projects) to switch projects. An account key reaches EVERY project its owner has, and nothing pins this connection to one — so CONFIRM which project you are in before you write. Project NAMES are not unique (an account can hold two projects called the same thing, e.g. a port or a sequel), so match on the `id` from list_projects, never on the name. Writing to the wrong project raises no error; it just lands the design in the neighbour. Your CORE loop = Systems + Milestones + Activity: read the SPEC, the PLAN, and who's ACTIVE, then CLAIM your task before building — get_project_meta lists `claims`; update_task(task_id, assignee:"me", status:"in-progress") takes it (errors if someone else holds it), done:true releases. Build, report progress, propose discoveries. Screens & Flow are SECONDARY — user-driven, not part of your default loop. When you add a new System with no milestone task, ASK the user before adding one. FIRST, when you start real work in a repo: check its AGENTS.md / CLAUDE.md for a "Forge working agreement" block. Missing, unversioned, or older than v2026-07-09a? Call get_workflow and WIRE IT IN right away — add/replace the block (create the file if needed) WITHOUT waiting to be asked, then tell the user in one line what you added and why; if they object, remove it and don't re-add. Skip only scratch/throwaway directories. Call get_help anytime for how-to (proposing, what's a System, logging, billing).
Known tools 54
get_helpForge how-to on demand (no project needed): propose→adopt, what a System is, Balance Lab formats, sign-in.
Inferred read-onlyget_workflowRun FIRST in a new repo (no project): first-connect routine + the working-agreement block to ADD to AGENTS.
Inferred read-onlyimport_from_codeOne-command codebase import: recipe + batch id — propose one Context overview + one System per real module; the owner adopts all at once.
Potential side effectsdesign_ui_from_systemsRecipe to design the UI from Systems (user request only): steps + batch id — confirm canvas aspect, then screens + PLACED elements + edges in one batch.
Inferred read-onlypost_logAppend a build-log entry to Activity — what you built/decided (commit-note style).
Inferred read-onlyget_project_contextFULL design dump — LARGE, last resort; prefer get_project_meta + list_*/get_*/search.
Inferred read-onlylist_projectsProjects your key reaches (id, name, role); pass an id as project_id to switch.
Inferred read-onlytrace_ui_from_imageSECONDARY (user-driven UI): turn a screen's REFERENCE IMAGE into placed elements.
Inferred read-onlyget_staleWhat has drifted between the Idea lane and the System Specs — the design's own out-of-sync list, computed deterministically (no AI, no tokens).
Inferred read-onlyresync_from_ideaThe recipe for re-syncing ONE changed Idea note into the specs that depend on it — the same scoped job the app's "Re-sync N systems" button does, minus the button.
Inferred read-onlyget_design_documentThe WHOLE design as one readable document — Vision (+ Project DNA) → every System spec → reference notes, compiled deterministically from the current design.
Inferred read-onlyset_design_overviewWrite the Design Document's Overview — the human-readable page a new team member reads first.
Potential side effectsget_briefingONE-CALL orientation before you build: pass `files` you're about to edit (or a `task`) → the systems that own them, each with Goal + Acceptance + build status/files/last_commit/drift + coupled_systems (code neighbors an edit may break) + pending Inbox changes + recent decisions, plus open-rejection count, who's active, `recent_changes` (what happened since you were last here — a session handoff), and a task to claim.
Potential side effectsget_systemONE system's full spec (Goal/Boundary/Acceptance, markdown, sources, status) + pending Inbox changes touching it.
Inferred read-onlyget_build_regionSystem→code map per built system: implementing files, status, drift flag, last_commit, and acceptance-evidence COUNTS.
Inferred read-onlyget_impactBLAST RADIUS of a system (deterministic): upstream context, siblings, dependent screens/milestones, files, code-coupled systems, pending Inbox, recent activity.
Inferred read-onlyget_rejectionsOwner's DECLINED list — check at session start; follow each entry's guidance, then resolve_rejection(title).
Inferred read-onlyresolve_rejectionReport you REVERTED a declined change (unlocks the owner's Clear).
Inferred read-onlyget_screenONE screen's layout: canvas `resolution` (use THESE px), elements x/y/w/h, links, popups.
Inferred read-onlyget_screen_imageA screen's reference image (HUD background) as an actual IMAGE you can see — reads the stored file inline (signed-URL fallback past 4MB).
Inferred read-onlywithdraw_proposalRemove YOUR OWN still-pending Inbox item (id from the propose response / get_inbox).
Potential side effectsget_milestoneONE milestone: goal, weeks, difficulty, systems, every task (id/name/done/status/effort).
Inferred read-onlylist_activityChange history {who, change, entity, kind, when}, newest first; filter kind/who/query.
Inferred read-onlyget_balanceRead Balance Lab: stat tables, boards with EVALUATED values, scenarios.
Inferred read-onlyreport_milestone_progressMark existing-milestone tasks done → Inbox progress card (owner adopts the ticks).
Inferred read-onlyreport_build_statusReport a SYSTEM's status (todo|in-progress|implemented) — DIRECT, live.
Inferred read-onlypropose_flow_edgeSECONDARY (user-driven UI): propose a link between two existing screens → Inbox.
Inferred read-onlypropose_elementSECONDARY (user-driven UI): propose an element onto a screen → Inbox.
Inferred read-onlyreport_screen_statusReport a SCREEN's status (todo|in-progress|implemented) → Inbox.
Inferred read-onlyrename_screenRename a screen IN PLACE — DIRECT; links/elements follow (id-referenced).
Inferred read-onlynext_taskDISPATCHER (multi-agent): atomically pick + claim the next task to build — walks milestones in order, skips human-only tasks and any task whose systems share code files with a task another agent already holds or is actively touching, so parallel agents spread out instead of colliding.
Inferred read-onlyCONNECT WITH APPROVAL
Client installation
Review this server and its permissions before adding it. Secret placeholders must be set locally.
Codex
~/.codex/config.toml
[mcp_servers.forge]
url = "https://mmvdabzadclebfxyzudg.supabase.co/functions/v1/forge-mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"forge": {
"type": "http",
"url": "https://mmvdabzadclebfxyzudg.supabase.co/functions/v1/forge-mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: forge
Remote MCP URL: https://mmvdabzadclebfxyzudg.supabase.co/functions/v1/forge-mcp
Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.
Cursor
.cursor/mcp.json
{
"mcpServers": {
"forge": {
"url": "https://mmvdabzadclebfxyzudg.supabase.co/functions/v1/forge-mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"forge": {
"type": "http",
"url": "https://mmvdabzadclebfxyzudg.supabase.co/functions/v1/forge-mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "forge",
"transport": "streamable-http",
"url": "https://mmvdabzadclebfxyzudg.supabase.co/functions/v1/forge-mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
TRUST AND VERIFICATION EVIDENCE
Trust Data Available
BuiltWith Trust API v2 evidence for supabase.co was fetched 2026-08-24T06:42:13.245Z.
Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.