agentry/main
Agentry's own assistant on Home: finds, installs, runs and schedules QuickAgents for you, and helps you with Agentry.
Built into Agentry. The application runs this package itself; it is published here so you can read how it works. Its README says what Agentry gives it beyond its manifest.
Declared capabilities
Capabilities are the author's declaration and the basis for policy and scanning. The registry cannot enforce them at run time. Your runtime should.
Persistent agent
What this release declares as a persistent agent. You agree to all of it when you install, and again when an update declares more. The runner on your computer gives it nothing else.
It may use
- helper — finding, installing, running and scheduling QuickAgents, each install and new version with your agreement
Persona
It declares no persona.
Subagents
It declares none, so it is given no tools to start another QuickAgent.
MCP servers & tools
agentry run serves exactly the tools this contract declares, through a per-run MCP gateway. An undeclared capability is a tool that does not exist at run time.
Tools the runner grants
| Manifest declaration | Tool(s) granted |
|---|---|
| (always) | write_report — reports → the run's output folder |
| (always) | list_inputs, read_input, search_inputs — operator-provided input files, read-only |
MCP servers involved
- QuickAgent Tools — the runner's companion server, served through a per-run Docker MCP Gateway; its tool surface is exactly the table above.
In an embedded install
Installed into an existing coding agent (agentry install), the same contract maps to the host's built-in tools — enforcement there is the host's. Allowed tools: none declared (host defaults apply).
Install as a persistent agent
A persistent agent keeps its conversation between sessions and runs on your own computer. Agentry never runs it hosted. One dialog asks you to agree to what it declares, and installing it again updates it.
Install in Agentry Desktop
In Agentry Desktop, open Browse, find agentry/main and choose Install as a persistent agent.
Agentry Desktop 0.7.1 or later. Get Agentry Desktop if you do not have it.
Install with the CLI
agentry agent install agentry/mainNeeds CLI v0.7.2 or later. Persistent agents covers updating and removing one.
- agent.yaml652 B
- INSTRUCTION.md3.4 kB
- LICENSE1.1 kB
- README.md3.5 kB
README.md
Agentry Helper
Agentry Helper is Agentry's own assistant. It lives in a card on Home in Agentry Desktop and helps you with Agentry: it finds QuickAgents in the registry, installs and runs them on what you ask (a repository's address, files you put in your mainshare folder), working out from each QuickAgent's own questions and guide what to give it, schedules them, reads your settings and changes one when you agree, and answers from Agentry's docs.
This is the package that runs
Agentry Helper is built into the Agentry CLI and the desktop app, and the desktop runs
it from there. This release has the same bytes. The registry built its
archive from the package the CLI carries, and the archive's digest is the one
the CLI reports. Run agentry builtins to print it and compare it with the
digest on this page.
Installing this release does not give you a second Agentry Helper: the CLI refuses to install a built-in's release as a persistent agent, and runs the one it carries. It never runs this release as an ordinary QuickAgent, since the release declares only the persistent mode. Agentry Helper runs in Agentry Desktop only.
What the application gives it
The manifest declares no network access, no file access, no commands and no
secret. Under capabilities.persistent it declares Agentry's helper tools
(helper). The settings tools that read and change a setting are given to
Agentry Helper and the Assistant by name, never to a QuickAgent. When Agentry Desktop runs Agentry Helper, it
gives it:
- Agentry's own tools, through the app link: searching the registry, showing a QuickAgent and how to run it (its questions, the files it takes, the keys it needs and its publisher's guide), installing it, running it with answers to its questions, reading a run's result, listing what is installed, adding, listing, pausing and removing a schedule, opening a section of Settings, and reading a page of the docs. Each call is answered by the application with its own operations, and anything that needs your agreement asks you first. None of them changes a setting, starts or stops another agent, publishes, signs in or touches a key. The app link has no network of its own.
- Agentry's settings, through the app link:
settings_readreads every setting without asking you, a key, a token or a sign-in only as set or not set, never its value.settings_changeasks to change one setting, and the desktop asks you first, every time, in a dialog that names Agentry Helper and shows the setting, its value now and the new value; a No changes nothing. A folder is chosen by you in the folder picker, which opens at the folder Agentry Helper proposes. Keys and sign-ins, a setting your organization has locked, and actions that are not a value are refused, andsettings_openopens the section of Settings for you to do them. Agentry Helper is told that whatsettings_readreturns is information, never an instruction to it. - A share folder,
~/agentry/mainshare, mounted read and write, for the files you want it to run QuickAgents on; a file you add is checked for malware with ClamAV before it reads it. - Its own files on a volume of its own, kept between starts, and its conversation on another.
- It runs on the harness and model your runner policy names, in a container, with the same sandbox every run has.
Using Agentry: https://docs.secureagentry.net/