Keelen

Official33 tools

by jamie7893

Autonomous dev team steered from chat: plain-English requests in, tested merged PRs out.

keelen MCP Server (io.github.jamie7893/keelen)

The keelen MCP server (io.github.jamie7893/keelen) provides an autonomous dev-team workflow driven from chat. It uses plain-English requests to run a loop that moves work from a PM-like stage to development and verification, producing tested, merged pull requests. The repo is presented as “keelen-mcp”.

🛠️ Key Features

Autonomous dev team steered from chat
Plain-English requests in
Tested merged pull requests out
MCP-focused server package (mcp, mcp-server)

🚀 Use Cases

Onboard an autonomous dev team “from your chat window”
Register the MCP server and connect a GitHub repo
Submit build requests that result in merged, tested PRs

⚡ Developer Benefits

Chat-driven control without a dashboard (“No dashboard required”)
Setup via a referenced SETUP.md for MCP-capable agents
Tool coverage indicated by 33 tools

⚠️ Limitations

Requires an MCP-capable agent and the ability to run the described setup flow
Workflow is framed around producing tested, merged pull requests rather than other outputs

Topics

ai-agentsautonomous-agentsclaudecodexdevtoolsglmkimimcpmcp-serverdeveloper-tools

Captured live from the server via tools/list.

list_projects

List the caller's keelen-saas projects.

Parameters

No parameters.

Raw schema

{
  "type": "object",
  "properties": {},
  "title": "list_projectsArguments"
}

create_project

Create a brand-new from-scratch project (no Chrome onboarding needed). Use this ONLY when the user wants a NEW repo scaffolded. If they already have a repo, use import_project(repo_full_name) instead — this tool would create a second, empty one beside theirs (list_github_repos() browses what the workspace can see). Scaffolds a new GitHub repo, a bootstrap-mode project, and submits `build_description` as the project's first Roadmap Request. `name` is a concise GitHub short repo slug (no owner); `project_kind` is REQUIRED and one of library | node_library | python_library | service | cli | web_app | godot_game | roblox_game; `preview_command` is required iff `project_kind == 'web_app'`. `engine` is OPTIONAL — one of claude_code | codex | glm | kimi (defaults to claude_code); codex, glm, and kimi require the workspace to have a matching connected credential. `org` is OPTIONAL — a GitHub organization login to create the repo inside (e.g. your company org); omit it to land the repo on a member's personal account. `private` defaults to True. `ci_runs_on` is OPTIONAL — the CI runner labels for the scaffolded workflow, e.g. ["self-hosted", "linux", "x64", "my-fleet"]. Omit it to inherit the workspace default (ubuntu-latest if unset). Labels no registered org runner carries are rejected, because GitHub would queue such a job forever rather than fail it. `framework` is OPTIONAL and `web_app`-only — one of vite | next (defaults to vite). It picks the scaffolded frontend rails: `vite` a vanilla-TypeScript SPA, `next` a Next.js app-router app. Passing it with any other `project_kind` is an error. The repo is created on the GitHub account of a workspace member with repo-create OAuth access (this path has no specific caller user), so the returned `repo` owner is whichever member's token resolved (or the chosen `org`). If no member has repo-create access — or the resolving member can't create in `org` — the call returns an actionable error. Returns {project_id, repo, thread_id, next_action, poll_after_seconds, next_step}; follow next_step (poll get_request_status with the returned thread_id). On the rare arm where the first Request failed to submit, next_action is "call_tool" with next_tool="submit_request".

Parameters9

name	string	required
build_description	string	required
project_kind	string	required
private	boolean	optional
preview_command	any	optional
org	any	optional
engine	any	optional
ci_runs_on	any	optional
framework	any	optional

Raw schema

{
  "type": "object",
  "properties": {
    "name": {
      "title": "Name",
      "type": "string"
    },
    "build_description": {
      "title": "Build Description",
      "type": "string"
    },
    "project_kind": {
      "title": "Project Kind",
      "type": "string"
    },
    "private": {
      "default": true,
      "title": "Private",
      "type": "boolean"
    },
    "preview_command": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Preview Command"
    },
    "org": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Org"
    },
    "engine": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Engine"
    },
    "ci_runs_on": {
      "anyOf": [
        {
          "items": {
            "type": "string"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Ci Runs On"
    },
    "framework": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Framework"
    }
  },
  "required": [
    "name",
    "build_description",
    "project_kind"
  ],
  "title": "create_projectArguments"
}

submit_request

Submit a free-form Roadmap Request to a project (enters PM intake). Submit ONE feature or intent per call; split a multi-feature ask into separate requests. `text` must be under 16000 characters. Returns {thread_id, status, next_action, poll_after_seconds, next_step}; follow next_step (re-check get_request_status after poll_after_seconds).

Parameters2

project_id	string	required
text	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "text": {
      "title": "Text",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "text"
  ],
  "title": "submit_requestArguments"
}

get_request_status

Get intake status + clarifying questions for a submitted request. Intake is async (~5min cadence) - poll periodically. Read `next_action`: "wait" (still processing), "answer_questions" (call answer_request with one answer per question), "done" (see generated_roadmap_item_ids), "cancelled" (terminal, no items), "failed" (see intake_failure_reason).

Parameters2

project_id	string	required
thread_id	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "thread_id": {
      "title": "Thread Id",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "thread_id"
  ],
  "title": "get_request_statusArguments"
}

set_product_vision

Set a project's product vision (prepended to PM/dev/QA iter prompts). `vision_md` is free-form markdown and must be non-empty. Tenant-scoped: a project not in the caller's workspace 404s. Returns {project_id, product_vision_md, updated_at, next_step}.

Parameters2

project_id	string	required
vision_md	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "vision_md": {
      "title": "Vision Md",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "vision_md"
  ],
  "title": "set_product_visionArguments"
}

set_product_goal

Set a project's product goal (prepended to PM/dev/QA iter prompts). `goal_md` is free-form markdown and must be non-empty. Tenant-scoped: a project not in the caller's workspace 404s. Returns {project_id, product_goal_md, updated_at, next_step}.

Parameters2

project_id	string	required
goal_md	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "goal_md": {
      "title": "Goal Md",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "goal_md"
  ],
  "title": "set_product_goalArguments"
}

list_roadmap

List a project's roadmap items (default: queued).

Parameters2

project_id	string	required
status	string	optional

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "status": {
      "default": "queued",
      "title": "Status",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "list_roadmapArguments"
}

reorder_roadmap

Reprioritise a project's QUEUED roadmap items to a new order. `ordered_ids` is the desired front-to-back order of queued roadmap-item ids (get them from list_roadmap). The first id becomes the highest priority — the cadence expands the lowest-priority_int queued item next. Horizon pins still dominate: a pinned-later item stays at the back and a pinned-now item at the front, regardless of position in `ordered_ids`. Ids that are unknown or no longer queued are skipped; duplicates are rejected. Returns {updated, queue, next_step}.

Parameters2

project_id	string	required
ordered_ids	array	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "ordered_ids": {
      "items": {
        "type": "string"
      },
      "title": "Ordered Ids",
      "type": "array"
    }
  },
  "required": [
    "project_id",
    "ordered_ids"
  ],
  "title": "reorder_roadmapArguments"
}

cancel_roadmap_item

Cancel / close a single roadmap item (get its id from list_roadmap). Use this to close a DELIVERED or duplicate item that keeps re-parking: when the work already shipped, every expand produces no dev-ready tasks and files a recurring `roadmap_item_parked` escalation you have to keep acking. Cancelling drops the item out of the expand queue AND resolves any open expand-lane escalation for it. Idempotent for an already expanded/cancelled item; refuses (409) while the item is actively being expanded (retry once that iter ends). Returns {id, status, changed, next_step}.

Parameters2

project_id	string	required
item_id	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "item_id": {
      "title": "Item Id",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "item_id"
  ],
  "title": "cancel_roadmap_itemArguments"
}

clear_horizon_pin

Clear a queued roadmap item's horizon pin (now/next/later → none). A horizon pin dominates the queue sort, so `reorder_roadmap` cannot move a pinned item out of its band — a stale `now` pin on a delivered/duplicate item clogs the front of the queue. This unpins it and reprices the queue so the item follows plain priority order again (and reorder_roadmap can then move it). Only queued items carry a settable pin (in-flight / shipped items derive theirs), so this refuses (422) on a non-queued item — use cancel_roadmap_item to close a delivered item. Returns {id, previous_pin, horizon_pin, next_step}.

Parameters2

project_id	string	required
item_id	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "item_id": {
      "title": "Item Id",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "item_id"
  ],
  "title": "clear_horizon_pinArguments"
}

rollback_roblox_place

Roll a Roblox project's place back to a previously-published version. For a `roblox_game` project, re-publishes the RETAINED build artifact for `version_number` (get published versions from the web Roblox Publishing card) — it never rebuilds from source, so rollback is fast + deterministic. This mints a NEW Roblox version pointing at the old build. Unknown version → 404; a version with no retained artifact → 422; a place open in Studio / rate-limited → 409 (retry); an invalid or unscoped Open Cloud key → 409 / 403. Returns {version_number, env, published_at, status, next_step}.

Parameters2

project_id	string	required
version_number	integer	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "version_number": {
      "title": "Version Number",
      "type": "integer"
    }
  },
  "required": [
    "project_id",
    "version_number"
  ],
  "title": "rollback_roblox_placeArguments"
}

project_status

Get a project's lifecycle status, open task count, iters today, pass rate, open escalations, pause state, freshness, and glm_peak_paused (the ephemeral GLM peak-hours skip that sets no pause columns — True means clean PRs hold + iters stop until glm_peak_resumes_at; an intentional cost gate, not a fault).

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "project_statusArguments"
}

list_escalations

List a project's OPEN escalations (the dead-ends to surface + resolve). `project_status` only COUNTS open escalations; this returns each one with its kind, reason, detail_md, recommended_action, and (when task-scoped) the blocked task's title + PR url. A "forever-paused" project with no open `project_pause` row is surfaced as a synthetic `orphan:<project_id>` row. Pass any returned `id` to resolve_escalation.

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "list_escalationsArguments"
}

resolve_escalation

Mark an escalation handled, clearing it from the queue. `decision_md` is a required short note (why/how it was resolved), appended to the escalation's detail_md as an audit trail. Resolving a project_pause / orphan_pause RESUMES the project (clears the pause); a task_block / operator_action is acknowledged only (the task keeps its board status). Idempotent. Accepts a real escalation UUID or a synthetic `orphan:<project_id>` id from list_escalations.

Parameters2

escalation_id	string	required
decision_md	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "escalation_id": {
      "title": "Escalation Id",
      "type": "string"
    },
    "decision_md": {
      "title": "Decision Md",
      "type": "string"
    }
  },
  "required": [
    "escalation_id",
    "decision_md"
  ],
  "title": "resolve_escalationArguments"
}

control_scheduler

Enable/disable, resume, or process-now a project's scheduler. `action` is one of: "enable" | "disable" | "resume" | "process_now". - enable/disable flip scheduler_enabled (the loop dispatches only enabled, status='active' projects). - resume clears a pause (peak/backoff/manual) so the project dispatches again. - process_now forces the next intake batch (no-op if the scheduler is off or no intake_pending request is queued). Returns the resulting scheduler state.

Parameters2

project_id	string	required
action	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "action": {
      "title": "Action",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "action"
  ],
  "title": "control_schedulerArguments"
}

archive_project

Archive a project (reversible shelve) — frees a project slot in-tool. Stops any in-flight machine, then flips the project to `archived`: it drops out of the per-tier project cap (freeing a slot for a new project) and the loop stops dispatching it, but the project + its history are kept and can be restored from the dashboard project page. Idempotent. Prefer this over `delete_project` unless you specifically want the project gone. Owner-scoped (an MCP key is owner-only); a project not in the workspace 404s.

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "archive_projectArguments"
}

delete_project

Soft-delete a project (the harder option) — frees a slot and hides it. Stops any in-flight machine, then flips the project to `deleted`: it disappears from `list_projects`, drops out of the project cap, and the loop stops dispatching it. The row is retained for audit but there is NO in-tool restore (unlike `archive_project`) — re-import the repo to reconnect it as a fresh project. Idempotent. Owner-scoped; a project not in the workspace 404s.

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "delete_projectArguments"
}

run_security_review

Kick off a deep, whole-repo security review of a project. Spawns a one-shot audit that scans the repo across a kind-aware taxonomy (secrets + git history, vulnerable/abandoned deps, injection, SSRF, path traversal, deserialization, crypto, info-leak, plus web authz/session/CORS, library API-misuse, game client-trust, or infra/CI as applicable) and posts findings to the project's Security review for human triage. You review the findings, then send the ones worth fixing into the loop as Requests; no fix is applied automatically. Billable; one audit in-flight per project. (Triggering is disabled while the feature is hardened for production: a project not on the operator allowlist — empty by default — returns a message instead of spawning; earlier results stay visible.) Requires a paid plan; a free or trial workspace gets a message telling the user to upgrade.

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "run_security_reviewArguments"
}

run_legal_exposure_review

Start a legal exposure review of a project's repository. Spawns a one-shot review machine that reads the checkout offline and maps what the code DOES onto commonly cited legal obligations, filtered by the project's saved compliance profile (jurisdictions plus eleven product facts). Findings land on the project's Legal page for a human to triage, and you read them with get_legal_exposure_findings. No change is ever applied automatically. THIS IS NOT LEGAL ADVICE AND IT IS NOT A LEGAL CLEARANCE. The result carries a `disclaimer_md` field: repeat it to the user before you summarise anything. The review is not exhaustive, so an empty result is never proof that anything is in order. Requires a saved compliance profile (409-shaped refusal without one), a plan that carries the feature, and the project on the operator allowlist. Billable; one review in flight per project. Every refusal comes back as `ok: false` with an actionable `next_step`, and starts nothing. Rate-limited per workspace.

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "run_legal_exposure_reviewArguments"
}

get_legal_exposure_findings

Read a project's legal exposure findings, worst exposure first. Returns each finding as an OBSERVATION plus the obligation commonly cited over that pattern, its citation, the date the citation was last checked, and whether counsel has reviewed the registry entry (`counsel_reviewed`, false today for every entry). `exposure_order` is an ORDER, never a score: there is no grade, no percentage, and no overall state in this output. Read `not_determinable` before you summarise: those are checks that could not reach a verdict, and leaving them out would turn a partial review into a clean answer. The `disclaimer_md` field must reach the user. `limit` defaults to 25 (max 100). `include_all` adds resolved, dismissed, and out-of-scope rows to the default actionable set. Read-only, no compute, no rate limit; it works even while the trigger is closed for the project.

Parameters3

project_id	string	required
limit	integer	optional
include_all	boolean	optional

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "limit": {
      "default": 25,
      "title": "Limit",
      "type": "integer"
    },
    "include_all": {
      "default": false,
      "title": "Include All",
      "type": "boolean"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "get_legal_exposure_findingsArguments"
}

run_control_gap_review

Start a source-bound security control gap review for one project. The review records bounded engineering observations under the user's selected Cyber Essentials or CMMC Level 1 or Level 2 context. It does not determine framework standing, and it does not make changes. The `disclaimer_md` field must be repeated to the user before any observation is summarised. The project needs a saved framework profile, an eligible plan, and a place on the operator allowlist. Billable; one review is allowed in flight per project. Refusals return `ok: false` with a next step and do not start work. Rate-limited per workspace.

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "run_control_gap_reviewArguments"
}

get_control_gap_findings

Read a project's security-control observations and review boundaries. The result groups records by evidence class without a total or an overall framework outcome. Read `not_determinable` and `outside_review_scope` before describing any observation. An empty group does not establish that a control is in place, and `disclaimer_md` must reach the user. `limit` defaults to 25 (max 100). `include_all` includes resolved, dismissed, and out-of-scope records. This read-only tool has no compute quota and remains available after the trigger closes for a project.

Parameters3

project_id	string	required
limit	integer	optional
include_all	boolean	optional

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "limit": {
      "default": 25,
      "title": "Limit",
      "type": "integer"
    },
    "include_all": {
      "default": false,
      "title": "Include All",
      "type": "boolean"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "get_control_gap_findingsArguments"
}

answer_request

Answer a thread's clarifying questions (status must be awaiting_answers). `answers` is a list of {"idx": <int from get_request_status>, "answer_md": <str, 1..2000 chars>}. Answer EVERY question exactly once. Flips the thread back to intake_pending; poll get_request_status again.

Parameters3

project_id	string	required
thread_id	string	required
answers	array	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "thread_id": {
      "title": "Thread Id",
      "type": "string"
    },
    "answers": {
      "items": {
        "additionalProperties": true,
        "type": "object"
      },
      "title": "Answers",
      "type": "array"
    }
  },
  "required": [
    "project_id",
    "thread_id",
    "answers"
  ],
  "title": "answer_requestArguments"
}

refine_request

Refine a done thread's generated roadmap items with feedback (max 5x). Status must be "done". `feedback_md` is 1..2000 chars. Moves the thread to refine_pending; poll get_request_status to see the revised items.

Parameters3

project_id	string	required
thread_id	string	required
feedback_md	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    },
    "thread_id": {
      "title": "Thread Id",
      "type": "string"
    },
    "feedback_md": {
      "title": "Feedback Md",
      "type": "string"
    }
  },
  "required": [
    "project_id",
    "thread_id",
    "feedback_md"
  ],
  "title": "refine_requestArguments"
}

signup

Create a Keelen account (or start agent login) — emails a 6-digit code. UNAUTHENTICATED — the only tool besides verify_email that works before a bearer key is configured. `email` is where the code is sent. Flow: signup(email) -> the user reads the 6-digit code from their inbox -> verify_email(email, code) returns a reveal-once API key -> save it as this server's `Authorization: Bearer <api_key>` header in your MCP client config -> reconnect -> get_onboarding_status() to continue setup. The code expires in 15 minutes; call signup again to resend. Response is uniform whether or not the email already has an account (enumeration-safe), so signup doubles as agent LOGIN. Rate-limited per IP and per email. ASK THE USER for `email` in chat and WAIT for their answer before calling this. Do NOT infer it from your client profile, the logged-in account, git config, or any other ambient source; if you already hold a candidate, echo it back and get an explicit yes first. Because this call doubles as LOGIN, a guessed address signs the user in to whatever workspace owns it, and the rest of setup then mints an API key on, and creates a project in, an account they did not choose.

Parameters1

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "email": {
      "title": "Email",
      "type": "string"
    }
  },
  "required": [
    "email"
  ],
  "title": "signupArguments"
}

verify_email

Redeem the emailed 6-digit code for a reveal-once workspace API key. UNAUTHENTICATED. `email` + `code` must match a code issued by signup(email) within the last 15 minutes (5 attempts max). The returned `api_key` is shown exactly ONCE — store it ONLY in the MCP client config ("Authorization: Bearer <api_key>"), NEVER in a repo or a file you might commit. Then reconnect this server with the header set and call get_onboarding_status(). An invalid/expired/consumed code returns a uniform error — call signup(email) for a fresh one.

Parameters2

email	string	required
code	string	required

Raw schema

{
  "type": "object",
  "properties": {
    "email": {
      "title": "Email",
      "type": "string"
    },
    "code": {
      "title": "Code",
      "type": "string"
    }
  },
  "required": [
    "email",
    "code"
  ],
  "title": "verify_emailArguments"
}

get_onboarding_status

Where onboarding stands + exactly what to do next (agent-readable). Reports engine_connected / github_connected / project_count / payment_status and a `next_action` string you should follow VERBATIM, polling this tool between steps. `next_action` is "call_tool" (call the tool named in `next_tool`, following `next_step`) until onboarding is complete, then "done". (`next_action_detail` echoes the pre-2026-07-29 dict shape and is DEPRECATED — it is removed 2026-10-29; read next_action/next_tool instead.) - engine step: send the user the dashboard /login link. Engine subscriptions (Claude / Codex / GLM) are connected in the DASHBOARD for security — NEVER ask for or paste engine credentials in this chat. - github step: call connect_github() for an install link. - project step: create_project(...) for a new repo, or import_project(repo_full_name) for an existing one. - launch step: poll get_provisioning_status(project_id) until ready. Re-checking is YOUR job — the server does not push. Also returns a `usage` block (pool / daily / machine-hours counters + tier caps) for capacity-aware automation clients.

Parameters

No parameters.

Raw schema

{
  "type": "object",
  "properties": {},
  "title": "get_onboarding_statusArguments"
}

open_dashboard

Get a one-click, pre-authenticated dashboard sign-in link for the owner. The engine-connect step — and any dashboard task (billing card update, a project page) — needs a signed-in browser. Because this server has already authenticated the workspace owner, this mints a single-use magic-link login token and returns a `/login?token=…` deep link: opening it signs the user straight into the dashboard (no email round-trip, no password) and lands them where onboarding left off. Send the user the returned `login_url`; it works once and expires in 15 minutes — call again for a fresh one.

Parameters

No parameters.

Raw schema

{
  "type": "object",
  "properties": {},
  "title": "open_dashboardArguments"
}

connect_github

Mint a GitHub App install link for the workspace owner. Returns an `install_url` — send it to the user to open in a browser. They pick the GitHub account/org, approve the install, and land on a "connected" page; then poll get_onboarding_status() until github_connected is true. The link expires in 10 minutes — call connect_github() again for a fresh one.

Parameters

No parameters.

Raw schema

{
  "type": "object",
  "properties": {},
  "title": "connect_githubArguments"
}

list_github_repos

List repos the workspace's GitHub connection can see (for import_project). Each entry has full_name, default_branch, private, language, pushed_at. Pass a `full_name` to import_project(repo_full_name) to connect it. Returns 409 if GitHub isn't connected yet — call connect_github() first.

Parameters

No parameters.

Raw schema

{
  "type": "object",
  "properties": {},
  "title": "list_github_reposArguments"
}

import_project

Connect an EXISTING GitHub repo as a Keelen project. This is the counterpart of create_project: use THIS tool when the user already has a repo, and create_project only to scaffold a brand-new one. `repo_full_name` is "owner/repo" — it MUST be visible to the workspace's GitHub connection (list_github_repos() to browse; a non-visible repo 404s). `engine` is OPTIONAL — one of claude_code | codex | glm | kimi (defaults to claude_code); codex/glm/kimi require a matching connected credential. `build_description` is OPTIONAL but STRONGLY recommended — a plain-language "what should Keelen build first?" submitted as the project's first Request so the loop has work; an imported project with no Request sits idle until you call submit_request(project_id, ...). `project_kind` is OPTIONAL — one of library | node_library | python_library | service | cli | web_app | godot_game | roblox_game | unknown. Omit it and the kind is auto-detected. PASS IT when the repo is a MONOREPO (apps in subdirectories), a stack with no standard root manifest (Java, Ruby, PHP, .NET, Elixir), or when you want a classification detection cannot infer — in those cases detection yields "unknown", which BLOCKS the dev lane until someone overrides it. A value you pass is authoritative and is never overwritten by later auto-detection. `stack` is the OPTIONAL language axis (python | node | rust | go | cpp) for a language-agnostic kind. `preview_command` is REQUIRED when project_kind is "web_app" (the command that serves the app locally, e.g. "npm run dev") and optional otherwise, where it overrides the detected one. Re-importing the same repo is idempotent (returns the existing project with already_exists=True). On a plan with no scheduled-project allowance the project is still created but with the loop OFF — next_step then steers to get_billing(). Otherwise follow next_step and poll get_provisioning_status(project_id).

Parameters6

repo_full_name	string	required
engine	any	optional
build_description	any	optional
project_kind	any	optional
stack	any	optional
preview_command	any	optional

Raw schema

{
  "type": "object",
  "properties": {
    "repo_full_name": {
      "title": "Repo Full Name",
      "type": "string"
    },
    "engine": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Engine"
    },
    "build_description": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Build Description"
    },
    "project_kind": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Project Kind"
    },
    "stack": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Stack"
    },
    "preview_command": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Preview Command"
    }
  },
  "required": [
    "repo_full_name"
  ],
  "title": "import_projectArguments"
}

get_provisioning_status

First-launch provisioning stages for a freshly created/imported project. Returns overall (provisioning | ready | errored), a 7-stage checklist, and a user-facing error_kind when a stage failed. next_action is "wait" with poll_after_seconds (~10s) while provisioning or errored, and "done" when overall is 'ready' (which also marks onboarding complete) — then steer the loop with submit_request(project_id, text). Tenant-scoped: a project not in the caller's workspace 404s.

Parameters1

project_id

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "project_id": {
      "title": "Project Id",
      "type": "string"
    }
  },
  "required": [
    "project_id"
  ],
  "title": "get_provisioning_statusArguments"
}

get_billing

Billing status + a Stripe checkout link when a NEW subscription is needed. `plan` is one of starter | pro | agency (default starter). When the workspace has NO live subscription and needs one (open-signup unpaid, churned, or converting from a free/trial tier), returns a `checkout_url` with next_action "browser" — send it to the user to open in a browser (the one setup step that can't happen in chat). Compute unlocks automatically once payment completes (a Stripe webhook flips the workspace to active); you do not need to block on it. A past_due workspace gets NO checkout — the fix is a card update in the dashboard billing page (a new checkout would create a second subscription); follow next_step. Subscribed or suspended-with-subscription states return checkout_url=None with an explanatory next_step.

Parameters1

plan

string

optional

Raw schema

{
  "type": "object",
  "properties": {
    "plan": {
      "default": "starter",
      "title": "Plan",
      "type": "string"
    }
  },
  "title": "get_billingArguments"
}

README.md10 KBView on GitHub

keelen-mcp

Steer Keelen, the AI coding agent for GitHub, from your chat.

Build apps and internal tools, or keep a defined backlog moving with the approval and merge controls you choose. Keelen turns requests into a standing roadmap, plans tasks, runs checks and opens pull requests in your repository.

Paste SETUP.md into a compatible MCP client and say "set me up." Your agent guides account setup and project creation. You confirm your email and complete AI credential connection, GitHub approval and any billing steps in the browser. Once connected, use chat to submit requests, set priorities and answer questions about the work.

Use your own supported AI subscription or API key with Claude Code, Codex, GLM, Kimi or Grok. Keelen adds no markup to AI usage. See how the MCP workflow works.

This repo is documentation only. It describes the hosted Model Context Protocol (MCP) server at https://keelen.ai/mcp. Keelen runs the coding workflow as a hosted service; this repository contains its documentation and client configuration examples.

Recorded Keelen MCP setup walkthrough

The recording shows an earlier setup flow. Follow the current instructions below for account verification and browser approvals.

Work you can hand over

Build an app or progress the fixes, tests and small features on a solo project's roadmap.
Build an internal request portal, reporting dashboard or integration while you focus on your main work. Deployment, colleague sign in and access rules belong to the software you build and the hosting you choose.
Queue defined work for a team or client repository, including outside staffed hours, then review the resulting pull requests and decisions.
Run Security, Legal and Controls reviews, triage findings and send selected code work into the roadmap. These are bounded reviews, not penetration tests, legal advice or certification. Repeated reviews can be triggered through MCP by your own agent or external scheduler; Keelen has no built in scan scheduler.

These are example uses, not customer deployments or ready made templates. You set priorities and choose plan approval, manual merge or other supported controls for each project. Keelen checks changes before they can merge and holds failed checks. Capacity, provider limits, missing information and decisions can pause progress; a queue is not a promise of finished work by a particular time.

AI engines and credentials

Connect credentials in the Keelen dashboard, never in the MCP chat. The workspace API key used by your MCP client is separate from these credentials.

AI engine	Supported connection
Claude Code	Claude Pro or Max sign in, or an Anthropic API key
Codex	ChatGPT sign in, or an OpenAI API key
GLM	Z.ai API key
Kimi	Kimi Code membership key or Moonshot Open Platform API key
Grok	SuperGrok sign in or xAI API key

Engine availability can differ by role. Grok here means the coding engine; it does not establish Grok Bot compatibility as an MCP client. OpenRouter is currently limited to Keelen's own test and benchmark projects and is not enabled for customer repositories.

Connect and set up

For a headless / API-key client (including Claude Code with a bearer key), connect the full surface without a token first, then hand your agent the setup instructions:

code

claude mcp add --transport http keelen https://keelen.ai/mcp

Then paste the contents of SETUP.md into your agent and say "set me up". It will ask for your email and the verification code that lands in your inbox. It then guides you through connecting an AI engine, approving GitHub access, creating or importing a project, and provisioning. Follow the dashboard links for the steps that require your browser.

The whole flow on one page, with the 50-second demo: https://keelen.ai/mcp-server/?utm_source=mcp-server&utm_medium=github-readme&utm_campaign=setup

Using a different client? See the per-client guides:

Claude (connector and Claude Code plugin)

Claude connects through OAuth — there is no API key to paste. Add the connector by URL and sign in through the browser consent page as the workspace owner:

Claude (web, Desktop, mobile) / MCP connector: https://keelen.ai/claude/mcp
Claude Code: claude mcp add --transport http keelen https://keelen.ai/claude/mcp

The OAuth grant requests read and write access by default; actions that need administrator authority ask for a step-up re-consent. Claude Code completes OAuth through a localhost callback on a randomly assigned port.

https://keelen.ai/claude/mcp is the Claude directory surface. It omits two tools, so it always exposes 40 tools: get_billing (billing is handled in the Keelen dashboard) and project_digest. The full surface, including both of those tools, is https://keelen.ai/mcp — use it for the headless / API-key path described above.

This repository is also the Claude Code plugin bundle. Two manifest pairs ship here and point at different surfaces:

Files	Ecosystem	MCP surface
`.claude-plugin/plugin.json` + `.mcp.json`	Claude Code plugin	`https://keelen.ai/claude/mcp`
`plugin.json` + `mcp.json`	agent-plugins.org	`https://keelen.ai/mcp`

The Claude Code plugin is installable from Anthropic's plugin directory once it is listed there; until then the per-client setup above is the supported path. Guides: Claude Code, Claude Desktop.

Building a Roblox game? Keelen can generate art, audio, meshes, skyboxes, and animated characters from plain-language requests — see roblox-assets.md.

Tools (abbreviated)

The server exposes 41 tools by default, or 42 tools when the default-OFF project_digest flag is enabled. project_digest is off by default and is not generally available. Two work with no key at all (signup, verify_email); the rest need the bearer key verify_email gives you. Full reference with parameters and behavior: TOOLS.md.

Group	Tools
Getting started	`signup`, `verify_email`
Onboarding	`get_onboarding_status`, `open_dashboard`, `connect_github`, `list_github_repos`, `import_project`, `get_provisioning_status`, `get_billing`
Projects & roadmap	`project_digest`, `list_projects`, `create_project`, `archive_project`, `delete_project`, `submit_request`, `get_request_status`, `answer_request`, `refine_request`, `set_product_vision`, `set_product_goal`, `get_product_vision`, `get_product_goal`, `list_roadmap`, `reorder_roadmap`, `cancel_roadmap_item`, `clear_horizon_pin`, `project_status`, `list_escalations`, `resolve_escalation`, `retry_blocked_task`, `rearm_roadmap_item`, `resolve_platform_policy_conflict`, `replan_task`, `close_task`, `control_scheduler`, `set_ui_review_scenario_cap`, `rollback_roblox_place`
Reviews	`run_security_review`, `run_legal_exposure_review`, `get_legal_exposure_findings`, `run_control_gap_review`, `get_control_gap_findings`

The server speaks MCP protocol revision 2026-07-28 as well as the earlier handshake revisions, from the same endpoint — no client action is required either way. Details in clients/generic.md; notable contract changes are recorded in CHANGELOG.md.

Portfolio Manager 0.4.1 is the manual template for managing selected existing public or private projects. Open the published 0.4.1 template and follow the guide to connect your own account. The separate Portfolio OAuth surface is available. The portfolio reference documents its scoped interface and optional 0.4 management features. Delegated actions require additional owner opt-in and are not enabled by copying the manual template; manual management reads use the selected-project consent. The primary MCP setup above is unchanged.

FAQ

Is my code touched before I agree to anything? No. Connecting the MCP server tokenless only lets you sign up and verify your email — it can't see, read, or modify any repository. connect_github requires an explicit GitHub App install you approve in your browser, and every following build runs inside an isolated per-tenant sandbox scoped to your own workspace.

What does it cost? Signing up, connecting GitHub, and creating a project are all free. Actually running the autonomous loop (compute) requires an active subscription — get_billing() returns a Stripe checkout link the moment you're ready. A free/unpaid workspace can do everything up through project creation. Work can start once billing is active, the project's loop is enabled and its setup requirements are satisfied.

How do I revoke access? Every key minted through this flow (or the dashboard) can be individually revoked. Go to your Keelen dashboard → Settings → API keys and delete the key. Revoking takes effect immediately; your agent will need a fresh key (rerun the signup flow, or mint a new one from the dashboard) to keep calling the server.

See SECURITY.md for the full security model: what the MCP server can and can't reach, rate limits, and key handling.

Support

Questions or issues: support@keelen.ai

Install

Tools (33)

list_projectscreate_projectsubmit_requestget_request_statusset_product_visionset_product_goallist_roadmapreorder_roadmap

Remote endpoint

Streamable HTTP

Hosted server - connect over the network, no local install.

url

https://keelen.ai/mcp

Add to VS Code Add to Cursor

claude_desktop_config.json

json

{
  "mcpServers": {
    "keelen": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://keelen.ai/mcp"
      ]
    }
  }
}

Desktop config is stdio-only; this bridges via mcp-remote. Native remote: Settings > Connectors.

keelen MCP Server (io.github.jamie7893/keelen)

🛠️ Key Features

🚀 Use Cases

⚡ Developer Benefits

⚠️ Limitations

Topics

list_projects

create_project

submit_request

get_request_status

set_product_vision

set_product_goal

list_roadmap

reorder_roadmap

cancel_roadmap_item

clear_horizon_pin

rollback_roblox_place

project_status

list_escalations

resolve_escalation

control_scheduler

archive_project

delete_project

run_security_review

run_legal_exposure_review

get_legal_exposure_findings

run_control_gap_review

get_control_gap_findings

answer_request

refine_request

signup

verify_email

get_onboarding_status

open_dashboard

connect_github

list_github_repos

import_project

get_provisioning_status

get_billing

keelen-mcp

Work you can hand over

AI engines and credentials

Connect and set up

Claude (connector and Claude Code plugin)

Tools (abbreviated)

FAQ

Support

Install

Remote endpoint

Related servers