CLEO MCP Server

io.github.kryptobaseddevPublished by GitHub account kryptobaseddevv2026.3.35Updated Oct 24, 2026

Task management for AI coding agents — CQRS 2-gateway pattern, 200 ops across 10 domains

Overview

AI-generated overview

CLEO is a local task-orchestration server that gives coding agents persistent tasks, sessions, memory, and a release lifecycle across projects.

What you can do
Create, search, show, complete and link hierarchical tasks with dependencies and priorities, including batch creation and readiness checks. Start, resume, checkpoint and brief work sessions so context survives between conversations. Store observations, patterns and sticky notes with semantic search and cross-session memory retrieval. Inspect code symbols and cross-project graphs, and coordinate multiple agents through waves, consensus and messaging. Run work through a staged lifecycle from research to release, with validation and integrity checks.
When to use it
Use it when you run one or more AI coding agents on long or multi-session projects and keep losing context between sessions. It suits teams that want structured task tracking, shared memory, and a defined lifecycle rather than ad-hoc prompts. It is less useful for small one-off edits.
Requirements
Runs locally over stdio as the npm package @cleocode/cleo. Node.js 24 or newer is required. The CLEO_PROJECT_DIR environment variable must point to a project directory that contains a .cleo folder, created by initializing the project with the CLI. No authentication is declared.
Before you install
It writes to project state: tasks, sessions, memory, and lifecycle records are created and changed, and some operations are mutations. It expects a project directory containing a .cleo folder, so point CLEO_PROJECT_DIR at the right project. The README describes agent spawning, multi-agent messaging, and release/promote steps, which can act on your repository.

Installation

In SourceWeft

  1. Open CLEO MCP Server in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

[CLEO — Agent First Task Orchestration]

CLEO

[npm version] [CI] [License: MIT] [Node.js Version]

[ko-fi]

CLEO is the work, evidence, memory and docs layer for AI coding agents. Agents (and you) drive it through one CLI, cleo <command>, and every call returns a single JSON envelope on stdout, so any agent harness can read it. CLEO keeps each change traceable from task to pull request to release, demands evidence before a task counts as done, remembers what was learned across sessions, and keeps documentation from rotting.

What it does

AreaWhat you getStart with
TasksSagas, epics and tasks with acceptance criteria, dependencies and ready waves; sessions that survive interruptionscleo add, cleo next, cleo focus <id>
EvidenceA task is done only when its gates are proven: commits and files, CI or test runs, lint and typecheck, linked to each acceptance criterioncleo done <id> --plan
Pull requestsA PR names its tasks (Closes: T#### AC1,AC2); an independent reviewer records a verdict per criterion; merging on green CI closes the taskcleo done --from-merge <pr> --dry-run, cleo review packet
DocsOne canonical home per doc, stable names, published mirrors and drift checks (see below)cleo docs query, cleo docs update
Memory (BRAIN)Observations, decisions and lessons kept across sessions and searchablecleo memory observe, cleo memory find, cleo briefing
Code intelligence (Nexus)A symbol graph of your repo: callers, blast radius, impact before you editcleo nexus analyze, cleo nexus impact <symbol>
ReleaseOne command tags the newest commit on main with green CI and publishes from changesetscleo release cut <version> --dry-run

Run cleo <command> --help for the full flags of any command.

Quick start

bash
npm install -g @cleocode/cleo
cd your-projectcleo initcleo session start --scope global --name "first session"
# Work is a tree: a saga at the root, epics under it, tasks under those.cleo saga create --title "Ship the auth hardening" --acceptance "login is rate limited|sessions expire"cleo add "Auth hardening" --type epic --parent T001 --acceptance "rate limit shipped|session expiry shipped"cleo add "Rate-limit the login endpoint" --parent T002 \  --acceptance "returns 429 after 5 failed attempts|a test covers the limit"
cleo next          # what to work oncleo focus T003    # identity, blockers, docs and memory for one task

Then do the work and close it.

  • With a pull request (the normal path): work on a branch, open a PR whose body says Closes: T003 AC1,AC2, check it with cleo done --from-merge <pr> --dry-run, get a per-criterion review (cleo review packet <pr> --task T003, then cleo review verdict), and merge on green CI. The merge sweep records the evidence and closes the task.
  • Without a pull request: cleo done T003 --plan names the evidence each gate needs; record it with cleo verify, then cleo complete T003.

Finish with cleo memory observe "<what you learned>" --title "<short title>" and cleo session end --note "<handoff>". The next session starts with cleo briefing.

Docs that don't rot

AI agents write a lot of documentation, and most of it rots. An agent that cannot find the existing doc writes a new one; names carry task ids and dates; two copies drift apart; nobody can tell which one is current. After a few months the docs folder is a pile of near-duplicates that agents keep extending.

CLEO Docs fixes this at the source:

  • One canonical home per doc. The store is the source of truth; files in docs/ are published mirrors. Find before you write (cleo docs query "<topic>"), and update the existing doc (cleo docs update <slug>) instead of adding a second one. cleo docs add is for genuinely new docs.
  • Stable names. Every doc name and slug is lowercase kebab-case with no task id, session id or date; task and decision links live in metadata. A CI gate (cleo check canon docs) refuses new violators.
  • One kind per folder. docs/adr, docs/spec, docs/plan, docs/research, docs/guide, docs/concept, docs/reference, docs/migration, docs/release and docs/note.
  • Drift is visible. cleo docs status compares each published mirror with the store, cleo docs rot flags docs whose covered code changed after them, and cleo docs doctor repairs the store.
  • Messy folders can be migrated. cleo docs modernize <folder> plans a kind, a stable name and keep/rename/merge/delete for every file, and applies it with a receipt you can roll back.

The how-to is docs/guide/cleo-documentation-sop.md.

Packages

The CLI is @cleocode/cleo; it depends on the rest. The main packages:

PackagePurpose
@cleocode/cleoThe cleo command-line interface
@cleocode/coreThe SDK: tasks, sessions, evidence, docs, memory, orchestration, release
@cleocode/contractsShared types and operation contracts
@cleocode/lafsThe JSON envelope format every call returns
@cleocode/nexusCode intelligence and the cross-project registry
@cleocode/caampProvider registry and agent package management
@cleocode/adaptersProvider adapters (Claude Code, Codex, Cursor, Gemini, Kimi, OpenCode and more)
@cleocode/skillsThe ct-* skills that teach agents how to use CLEO
@cleocode/cantThe CANT agent-definition language

Development

Requires Node.js >= 24.16.0 and pnpm.

bash
git clone https://github.com/kryptobaseddev/cleo.gitcd cleopnpm installpnpm run build

Repository rules (package boundaries, architecture gates, evidence) are in AGENTS.md; see CONTRIBUTING.md to contribute. Every change lands through a pull request that names its tasks; releases are cut from main with cleo release cut.

The story of CLEO

CLEO began as one developer's refusal to keep losing the thread: powerful agents that forgot too easily, ambitious projects that died on the shelf. The Founding Story tells it from the builder's side, and The Awakening Story from CLEO's.

Support

If CLEO helps your workflow, consider supporting development:

[ko-fi]

License

MIT License - see LICENSE for details.

Star History

[Star History Chart]

Source: README.md at commit cbfc34e

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v2026.3.35LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts