FEATURES

43 functions. One cockpit.

Everything Maegor does today, grouped by where you meet it. Nothing on this page is a plan: if it is listed here it is built, and it is covered by tests.

functions: 43
MCP tools: 28
squads: 14
areas: 7

The cockpit

Where you watch the work happen.

Multi-pane terminal

Concurrent agents, each in its own PTY and its own OS process group, each with a live terminal you can type into.

PTY + pgid

Panes you arrange

Split, resize and rearrange. The arrangement is a tree saved per project, so it outlives the mission that created it.

split · resize · persist

Console

A second home that outlives every mission: one pane per agent, a conversation you pick up rather than a process you restart.

Workspace

A pane is a container, not a terminal slot. Terminal, editor, file explorer and browser preview share the same grid.

terminal · editor · files · browser

Editor and files

Monaco over the project tree, sandboxed to the project and resolving symlinks. A file under an active write lease opens read-only and says which agent holds it.

monaco

Browser preview

Finds the dev server URL in an agent's output and docks it beside the terminal, at desktop, tablet or phone width. It never splits your grid on a guess.

localhost:*

Command palette

Every action in the app reachable without leaving the keyboard.

Ctrl+K

Missions

A brief becomes a graph, and the graph becomes work.

Task graph

Dependencies and parallel groups, with cycles and dangling dependencies refused at plan time, before a single agent is dispatched.

0 cycles

Three ways to run

Free drives one agent yourself, Squad runs a declarative team, Agentic lets Maegor decompose and delegate. Same core, different amount of autonomy.

free · squad · agentic

Clarifying questions

The orchestrator asks what it needs before it plans, and your answers become part of the brief. Where an answer contradicts an assumption, the answer wins.

pre-plan

Review center

Agents submit reviews and findings through MCP. A rejected review sends the task back instead of quietly integrating it.

review_submit

Gates and diffs

The real git diff of an agent's worktree, next to the gate that judged it.

git diff

Artifacts

The reports, specs and contracts an agent produced, addressed by mission and task and readable long after the session ended.

artifact_create

The squad

Who does the work, and what each one is allowed to be good at.

Squads

Declarative YAML teams: roles, models, file globs, gates and budgets. Fourteen ship with the app, from a one-agent Solo to a twelve-agent web build.

14 YAML

Squad editor

Write a squad inside the app and have it validated on save, with the offending field named.

validate on save

Agents

A CLI, a model, a set of skills and an effort level composed into one specialist you can dispatch again.

cli · model · skills · effort

Arsenal

The roster of specialists a project has built up, reusable across its missions.

roster

Skills

An installable capability that attaches to an agent and takes effect on its next run.

install → next run

Providers

The agents you already pay for, driven from one place.

Agent adapters

Claude Code, Codex and OpenCode drive the binaries you already have. Detection executes each one, it never trusts a config file, and every flag was read from a local help output.

--help verified

Live sessions

A headless agent that stays alive between turns: the next prompt is a line written to a stdin that never closed, so there is no cold start and no lost context.

stream-json

Native agent

Maegor's own tool loop, for the work where there is no third-party CLI worth driving.

adapter: native

Bare shell

A pane with no agent in it at all, in the same grid and under the same kill switch.

Honest event mapping

Headless output becomes typed events, and the pane says so when the mapping degrades instead of pretending it understood.

typed events

Enforcement

Checks in the runtime, not lines in a prompt.

Worktree per agent

A git worktree and a branch of its own for every executing agent. An agent's writes never reach your working tree.

1 branch / agent

File leases

Two agents that could write the same file are refused at dispatch, by glob intersection, not when the merge falls apart.

glob ∩

Quality gates

Gates run inside the agent's worktree against the real git diff. A failing required gate blocks integration.

git diff

Kill switch

Terminates whole process groups, grandchildren included. A test spawns a grandchild and asserts that it dies.

kill -pgid

Pause and resume

SIGSTOP and SIGCONT, per agent or per whole mission, without losing the terminal.

SIGSTOP · SIGCONT

Approvals

Destructive commands, out-of-scope writes and budget overruns stop the agent until you decide. An expired request is a refusal, never an approval.

queue + TTL

Watchdogs

A wall-clock ceiling and a heartbeat timeout catch a hung agent, terminate it, and record why.

wall clock + heartbeat

Restart safety

Sessions the database still believes are running are marked orphaned when the project opens. Nothing is auto-resumed after a crash.

orphaned

Sandbox and secrets

Tokenised destructive-command detection, path sandboxing that resolves symlinks, and secret redaction on the persistence path. Modes change how often you are asked, never whether a check runs.

paths · secrets · rm -rf

Budgets

A ceiling per mission. Crossing it stops the agent and asks you, it does not quietly keep spending.

ceiling

Memory and data

What a project remembers between missions.

Three-level memory

Mission, project and agent. The project level is why the second mission on a repository is cheaper than the first.

mission · project · agent

Memory graph

Entries link to each other, so the shortest path between two of them is a query rather than a scroll.

memory_path

Context budget

The builder picks what is relevant under an explicit token ceiling, instead of pasting everything it has and hoping.

token ceiling

Event log

Typed and append-only over SQLite, searchable and filterable by agent, task, kind, severity and free text.

append-only

Cost telemetry

Live cost per agent, with its source labelled. When a provider reports nothing the answer is unknown, never $0.00.

Reported | unknown

Surfaces

The other ways in.

MCP server

Twenty-eight tools over missions, tasks, memory, artefacts, leases and gates, behind a per-mission token. Every tool checks the id belongs to the calling session.

28 tools

CLI

The maegor binary and the desktop app share one Rust domain core: same state machine, same database, same event log.

maegor

Dictation

An optional local speech service for describing a mission out loud. Nothing leaves the machine.

offline

Doctor

Checks the install, the detected adapters, and the worktrees a cancelled mission left behind.

maegor doctor

Credit-free demo

A deterministic fake agent with six behaviours runs the whole product with no network and no LLM spend.

6 behaviours

SQUADS

Declarative teams. Validated at load.

A squad is a YAML file: roles, models, file globs, gates, budgets. Break a field and it tells you which one, before anything runs. Fourteen ship with the app.

fullstack-factory.yaml7 agents

Full-stack factory

Implements a complete feature across backend and frontend, in parallel against a contract both sides treat as frozen.

discoveryarchitecturebackendfrontendtestsreview+1

web-development.yaml12 agents

Web development

Builds a whole web feature, discovery through review. Sequential on purpose, so every handoff is against what was actually delivered.

discoveryarchitectureinterface-specbackendfrontendintegration-tests+6

bug-hunt.yaml8 agents

Bug hunt

Scans the codebase by area in parallel, then triages, fixes and locks the findings.

surveyhunt-webhunt-backendhunt-opstriagefix+2

bug-investigation.yaml5 agents

Bug investigation

Reproduces a defect, fixes the root cause, and locks it with a test.

reproducediagnosefixregression-testreview

security-audit.yaml4 agents

Security audit

Finds vulnerabilities, secrets and risky dependencies, then fixes the findings that matter.

scan-codescan-depsfixreview

database-migration.yaml5 agents

Database migration

Plans, applies and verifies a schema move, with backfill and rollback safety.

planmigratebackfillverifyreview

feature-implementation.yaml6 agents

Feature implementation

Implements a feature across a Rust core and a React shell, regenerating the shared types in between.

discoveryarchitecturecoreshelltestsreview

test-coverage.yaml4 agents

Test coverage

Raises coverage without touching production code, gated on the real coverage delta.

analyseunit-testsintegration-testsreview

refactoring.yaml4 agents

Refactoring

Removes dead code and duplication with behaviour provably unchanged.

surveyrefactorregressionreview

api-design.yaml4 agents

API design

Designs an API contract and locks it with tests on both sides of it.

surveydesigncontract-testsreview

dependency-upgrade.yaml5 agents

Dependency upgrade

Upgrades dependencies in reviewable groups, verifying the tree after each one.

assessupgrade-jsupgrade-nativeverifyreview

performance-profiling.yaml4 agents

Performance profiling

Finds bottlenecks with evidence, optimises them, and proves the win with benchmarks.

profileoptimizebenchmarkreview

commit-organizer.yaml3 agents

Commit organizer

Splits a dirty working tree into one commit per logical change, each with a message that names it, then pushes.

splitverifypush

solo.yaml1 agents

Solo

One agent does the whole thing, for a brief too small for handoffs to be worth what they cost.

Your engineering team is waiting.

Open a repository, describe a mission, and watch four agents work in parallel, inside the rules you set.

No credits burned in the demo, a deterministic fake agent runs the whole suite.