Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/XxYouDeaDPunKxX/canon-boundary-guard-codex/llms.txt

Use this file to discover all available pages before exploring further.

Canon Boundary Guard is installed as a Codex plugin. Codex discovers it through the marketplace manifest at .agents/plugins/marketplace.json, which points to the plugin directory at plugins/canon-boundary-guard-codex/. Inside that directory, .codex-plugin/plugin.json identifies the plugin, declares its skills path, and points to the hook configuration. Once Codex loads the plugin, it makes the bundled canon-boundary-guard skill available to the session and — when plugin hooks are enabled — wires the inject_frame.py script into the PreToolUse lifecycle event for matched write tools.

​
Directory Tree

The full layout of the plugin within the repository is: 
.agents/
`-- plugins/
    `-- marketplace.json
plugins/
`-- canon-boundary-guard-codex/
    |-- .codex-plugin/
    |   `-- plugin.json
    |-- hooks/
    |   `-- hooks.json
    `-- skills/
        `-- canon-boundary-guard/
            |-- SKILL.md
            |-- agents/openai.yaml
            |-- references/frame.md
            `-- scripts/inject_frame.py

​
File Roles

FileRole
.agents/plugins/marketplace.jsonExposes the plugin as a marketplace entry so Codex can discover and install it
plugins/canon-boundary-guard-codex/.codex-plugin/plugin.jsonPlugin manifest: identifies the plugin, points to skills and hooks
skills/canon-boundary-guard/SKILL.mdFull session-level operating frame read by Codex on activation
skills/canon-boundary-guard/agents/openai.yamlCodex-facing skill metadata: display name, short description, default prompt, and policy
skills/canon-boundary-guard/references/frame.mdCompact classification frame injected by the hook before write tools
skills/canon-boundary-guard/scripts/inject_frame.pyReads frame.md and emits the PreToolUse hook payload as JSON
hooks/hooks.jsonWires inject_frame.py to the PreToolUse hook for matched write tools

​
plugin.json

The plugin manifest at plugins/canon-boundary-guard-codex/.codex-plugin/plugin.json identifies the plugin to Codex, provides display metadata, and declares the relative paths for skills and the hook configuration file. 
{
  "name": "canon-boundary-guard-codex",
  "version": "1.0.0",
  "description": "Codex plugin for keeping provenance and cognitive layers separate across a session.",
  "author": {
    "name": "XxYouDeaDPunKxX",
    "url": "https://github.com/XxYouDeaDPunKxX"
  },
  "homepage": "https://xxyoudeadpunkxx.github.io/canon-boundary-guard-codex/",
  "repository": "https://github.com/XxYouDeaDPunKxX/canon-boundary-guard-codex",
  "license": "CC-BY-SA-4.0",
  "keywords": [
    "codex",
    "skill",
    "plugin",
    "hooks",
    "provenance",
    "agent-control",
    "cognitive-layers",
    "ai-workflows"
  ],
  "skills": "./skills/",
  "hooks": "./hooks/hooks.json",
  "interface": {
    "displayName": "Canon Boundary Guard",
    "shortDescription": "Provenance boundary guard for Codex",
    "longDescription": "Canon Boundary Guard packages a Codex skill and optional lifecycle hook for keeping project evidence, chat context, operator instructions, working hypotheses, and model assumptions in separate cognitive layers.",
    "developerName": "XxYouDeaDPunKxX",
    "category": "Productivity",
    "capabilities": [
      "Workflow",
      "Automation"
    ],
    "websiteURL": "https://xxyoudeadpunkxx.github.io/canon-boundary-guard-codex/",
    "defaultPrompt": [
      "Use Canon Boundary Guard as the session operating frame.",
      "Use $canon-boundary-guard for provenance and cognitive-layer separation."
    ],
    "brandColor": "#536878"
  }
}
The brandColor field value #536878 is the canonical brand color for Canon Boundary Guard. It is used by Codex UIs that render plugin identity — such as the plugin picker and marketplace listings — to display the plugin with its designated slate-blue tone.

​
agents/openai.yaml

The agents/openai.yaml file provides Codex-facing skill metadata. It declares the display name, a short description used in skill lists, the default invocation prompt, and the policy that allows Codex to activate the skill implicitly when context suggests it is needed. 
interface:
  display_name: "Canon Boundary Guard"
  short_description: "Session provenance frame for Codex cognitive-layer separation"
  default_prompt: "Use $canon-boundary-guard as the session operating frame for provenance and cognitive-layer separation."

policy:
  allow_implicit_invocation: true

​
marketplace.json

The .agents/plugins/marketplace.json file exposes the plugin to the Codex marketplace discovery mechanism. It declares the plugin’s name, display name, source location, installation policy, and category. 
{
  "name": "canon-boundary-guard",
  "interface": {
    "displayName": "Canon Boundary Guard"
  },
  "plugins": [
    {
      "name": "canon-boundary-guard-codex",
      "source": {
        "source": "local",
        "path": "./plugins/canon-boundary-guard-codex"
      },
      "policy": {
        "installation": "AVAILABLE",
        "authentication": "ON_INSTALL"
      },
      "category": "Productivity"
    }
  ]
}

hooks.json

The hook wiring manifest that connects inject_frame.py to the PreToolUse lifecycle event.

inject_frame.py

The hook payload script that reads frame.md and emits structured JSON for Codex.

frame.md

The compact provenance classification frame injected before write tools.

SKILL.md

The full session-level operating frame read by Codex on activation.

Build docs developers (and LLMs) love

Get started for freeTalk to us