Backend Reference logo

Backend Reference

OrganizationPopular
agent0ai
Backend Reference

Read-only reference for the backend contracts that the frontend depends on, including routing, APIs, auth, layers, and module resolution.

Overview

Publisheragent0ai
Repositoryspace-agent
Skill nameBackend Reference
Stars
1.4K
Forks
323
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by agent0ai on GitHub. Read the source before you install it.

Installation

Install the Backend Reference AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/agent0ai/space-agent.git /tmp/space-agent
mkdir -p .claude/skills
cp -r /tmp/space-agent/app/L0/_all/mod/_core/skillset/ext/skills/development/backend-reference .claude/skills/agent0ai-backend-reference
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Backend Reference in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Backend Reference on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Backend Reference is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Use this skill when frontend work depends on understanding backend behavior. This skill is read-only context, not authorization to change backend files.

Hard Boundary

  • Do not edit server/, commands/, or packaging/ from this skill set.
  • If the task truly requires backend changes, stop and ask instead of treating this skill as permission.
  • Backend changes are non-standard in this project. Only ask for them when frontend-only work cannot safely enforce security, integrity, cross-user isolation, or runtime stability.
  • When you ask, explain why the frontend solution would be insufficient and describe the narrow backend change required.

Page And Request Flow

Current request order is:

  1. API preflight handling
  2. /api/proxy
  3. /api/<endpoint>
  4. /mod/...
  5. /~/... and /L0/..., /L1/..., /L2/...
  6. page shells and page actions

The browser app mounts through page shells in server/pages/. /login is public. / and /admin require authentication.

API Families

Important frontend-facing endpoint families are:

  • app files: file_list, file_paths, file_read, file_write, file_delete, file_copy, file_move, file_info, folder_download
  • local history: git_history_list, git_history_diff, git_history_preview, git_history_rollback, git_history_revert
  • modules: module_list, module_info, module_install, module_remove
  • runtime and identity: extensions_load, password_generate, password_change, user_crypto_session_key, user_self_info

These endpoints are thin wrappers over shared helpers in server/lib/customware/ and server/lib/auth/.

file_write still defaults to whole-file replacement, but it also supports operation: "append", "prepend", and "insert". Insert writes accept exactly one anchor through line, before, or after, use the first literal before or after match, treat line as a 1-based insertion point, and require utf8.

file_list and file_paths support access: "write" or writableOnly: true for writable-only discovery. They also support gitRepositories: true; with a pattern such as **/.git/, file_paths returns local-history owner roots like L1/team/ and L2/alice/ without exposing reserved .git metadata paths.

folder_download supports HEAD for permission-only validation and GET or POST for the streamed ZIP attachment. It creates the archive in server/tmp/ only for the actual download response, after the shared file-access layer approves the requested folder path.

git_history_list, git_history_diff, git_history_preview, git_history_rollback, and git_history_revert are enabled by CUSTOMWARE_GIT_HISTORY. The backend maps the requested path to a writable L1/<group>/ or L2/<user>/ owner repo, enforces permissions, paginates list metadata with optional file filters, reads per-file diffs separately, previews travel or revert affected files with optional operation-specific file patches, suppresses history scheduling during rollback, preserves ignored L2 auth files, preserves forward-travel refs during rollback when possible, and creates inverse commits for revert.

user_self_info is the frontend-facing identity snapshot. Frontend callers derive writable app roots from username, managedGroups, and _admin membership in groups without authorizing backend edits.

password_change is the authenticated self-service password-rotation endpoint for the current user. Frontend code should call it instead of writing ~/meta/password.json directly, because the backend validates the current password, seals the replacement verifier, clears sessions, and clears the current auth cookie.

user_crypto_session_key is the authenticated localStorage-restore endpoint. It returns the current session-derived wrapping key by hashing the live backend sessionId with the server-held session secret, so the browser can decrypt or encrypt the one persisted userCrypto localStorage blob without the server persisting that wrapping key or the user master key.

Module And Extension Resolution

  • /mod/... resolution goes through the layered customware model.
  • maxLayer constrains module and extension resolution and defaults to 2.
  • /admin effectively clamps module and extension resolution to L0.
  • Frontend HTML extensions resolve through ext/html/....
  • Frontend JS hooks resolve through ext/js/....
  • Frontend modules may also enumerate other extension-owned assets such as ext/panels/*.yaml through file_paths plus file_read when they only need readable logical file discovery and contents. extensions_load remains the backend contract for HTML and JS extension resolution, keeping maxLayer at the top level, ordered patterns groups in the request, and ordered grouped results with matching patterns plus resolved extensions.

Auth And User Storage

  • The server issues the space_session cookie and validates it by hashing it with a backend-held secret.
  • When the current login may auto-restore userCrypto on the same browser profile, the browser stores one encrypted userCrypto blob in localStorage, and authenticated browser code fetches the current session-derived wrapping key from /api/user_crypto_session_key.
  • User metadata lives at L2/<username>/user.yaml.
  • L2/<username>/meta/password.json stores a server-sealed SCRAM verifier envelope, not a self-sufficient plaintext verifier.
  • L2/<username>/meta/logins.json stores backend-keyed session verifiers plus signed session metadata.
  • Authenticated self-service password changes still stay backend-owned through /api/password_change; browser code may edit L2/<username>/user.yaml directly, but not the password verifier file.
  • User-owned modules live under L2/<username>/mod/.
  • Backend auth keys come from shared environment injection through SPACE_AUTH_PASSWORD_SEAL_KEY and SPACE_AUTH_SESSION_HMAC_KEY for multi-instance deployments, or from the local fallback server/data/auth_keys.json during single-instance development.

Why Frontend Developers Need This

  • to understand why a frontend file API call succeeds or fails
  • to understand where module and extension inheritance comes from
  • to understand why admin UI is firmware-clamped while normal app routing is layered
  • to understand the user and group storage model without editing backend logic
  • to identify the small set of cases where backend ownership is mandatory instead of merely convenient

Mandatory Doc Follow-Up

  • If backend contracts that the frontend depends on change, update the mirrored backend docs and the development skill subtree in the same session, even though this skill remains read-only guidance.

Frequently asked questions

What does the Backend Reference AI skill do?

Read-only reference for the backend contracts that the frontend depends on, including routing, APIs, auth, layers, and module resolution.

Why use Backend Reference on TypingMind?

Because you install it once and use it with any model. Backend Reference is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Backend Reference in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/agent0ai/space-agent/tree/main/app/L0/_all/mod/_core/skillset/ext/skills/development/backend-reference. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Backend Reference?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Backend Reference?

As many as you like. As long as a model supports skills, you can use Backend Reference with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Backend Reference AI skill free?

Yes. It is published on GitHub by agent0ai under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇