ASIMP :: Security Platform
ANSIBLE SYSTEM INTEGRITY MANAGEMENT PLATFORM // baselines & hardening
MODE:

Documentation

Home Gitbook & dsom summary index Asimp for ai agents: deep state of mind (dsom) protocol integration Ansible execution environments (ee) & sovereign gitea integration guide Ansible best practices & fqcn standards Ansible playbook and document architecture map Architecture & design Configuration & variables Debian gnu/linux hardening & audit guide Review & adoption of dsom ansible configuration guide (v3.6.2) DiΓ‘taxis documentation framework Deep state of mind (dsom) governance framework System architecture & security posture Operational recipes index Execute tool workflows Legal notice & disclaimer Asimp local testing matrix & telemetry spec Lynis auditing & playbook analysis Openscap integration & playbook analysis The enterprise linux openscap operational guide: automated vulnerability auditing, baseline compliance, and safeguarded air gapped remediation Opensuse / suse linux enterprise hardening & audit guide Rootless podman 5+ & quadlet orchestration Asimp standard footer patcher reference Google okf frontmatter patcher reference Ansible config parameters reference Component & tool index Sovereign feedback collector reference Llm xml context compiler reference Role lynis ansible reference Mock scanning utility reference Openwiki emulator specification Sovereign os role patcher reference System core playbooks reference Jekyll pre processor reference Role reporting asimp reference Suse/sled sysctl hardening role reference Role update ubuntu asimp reference Sitemap & link integrity validator reference Enterprise linux cis level 2 audit & hardening guide Local knowledge first & metadata discovery Troubleshooting & fallback behaviors Quickstart onboarding guide Ubuntu 24.04/26.04 lts hardening & audit guide


About ASIMP

ASIMP is an automated framework built with Ansible to assess host compliance, configure OS parameters, secure SSH, and verify integrity.

ASIMP Local Testing Matrix & Telemetry Spec

Ansible >= 9.0.0 OpenSCAP: CIS L2 Lynis: Enabled debsums: Enabled

ASIMP Local Testing Matrix & Telemetry Specification

This document defines the architectural specification and implementation protocol for the local multi-OS testing matrix and bidirectional feedback telemetry pipeline within the Ansible System Integrity Management Platform (ASIMP).


1. Architecture & Mode Separation Protocol

To ensure a strict boundary between testing telemetry and user execution, ASIMP employs a conditional mode separation protocol controlled by the environment variable EXECUTION_MODE and the Ansible inventory variable execution_mode.

                        +----------------------------------------+
                        |  Execution Context Init (Ansible/Bash) |
                        +----------------------------------------+
                                            |
                                            v
                                 Is EXECUTION_MODE=dev?
                                 (or execution_mode == 'dev')
                                   /                  \
                                  /                    \
                                 v                      v
                      [ Yes: Developer Mode ]     [ No: User Mode ]
                                 |                      |
                      +----------------------+  +-------------------------+
                      | Enable Telemetry     |  | Zero Telemetry          |
                      | Gather Debug Metrics |  | Lightweight Execution   |
                      | Hook API/Jules CLI   |  | Clean Exit / Zero Hooks |
                      +----------------------+  +-------------------------+

Modes Defined

  1. Developer / Feedback Mode (EXECUTION_MODE=dev / execution_mode=dev):
    • Telemetry Collection: Full capture of tasks, outcomes, stdout/stderr, kernel state, exit codes, and diff analysis.
    • Dynamic Feedback: Automatic packaging of telemetry into /tmp/jules_telemetry.json and streaming back to Google Jules and active GitHub Pull Requests.
    • Performance Profiling: Captures resources utilized by Podman containers.
  2. User / Production Mode (EXECUTION_MODE=user / execution_mode=user):
    • Zero Telemetry: All external feedback hooks, API dispatches, and temporary telemetry file writing are physically bypassed.
    • Lightweight Execution: Minimum resource overhead, executing native ASIMP baselining and hardening strictly within host borders.

2. Multi-OS Matrix Orchestration Architecture

Execution runs natively inside Windows WSL2 (Ubuntu 26.04 LTS) and manages a multi-distribution Podman 5+ container matrix.

Container Grid

  • Ubuntu 24.04 LTS (ubuntu24): Active baseline testing.
  • Ubuntu 26.04 LTS (ubuntu26): Bleeding-edge environment validation.
  • AlmaLinux 9 (alma9): RHEL-compatible ecosystem validation.
  • Debian 12 (debian12): Pure Debian system compliance validation.

Failure Diagnostics & Capture

Inside playbooks/matrix_test.yml, task-level error handlers (block/rescue/always) capture failure states:

  • block: Contains the target tests (ASIMP scans, compliance audits, shell commands).
  • rescue: Activates upon task failure. Registers failure messages, step location, system state, and appends them to host-level telemetry arrays.
  • always: Executes regardless of outcome, ensuring the feedback_collector role is loaded to preserve and structure the telemetry data.

3. Bidirectional Jules CLI & GitHub PR Bridge

The bridge script (scripts/jules_gh_feedback.sh) acts as an idempotent dispatch courier between the local testing matrix, Google Jules CLI, and the GitHub Pull Request API.

       +----------------------------+
       |   /tmp/jules_telemetry.json|
       +----------------------------+
                      |
                      v
         +-------------------------+
         | scripts/jules_gh_feedback.sh
         +-------------------------+
               /                 \
              v                   v
     +-----------------+   +-------------------------+
     | Google Jules CLI|   | GitHub API / gh CLI     |
     | (jules chat)    |   | (gh pr comment <PR_ID>) |
     +-----------------+   +-------------------------+

Integrations and Fallbacks

  • Google Jules CLI: Automatically feeds markdown-formatted test summaries and detailed bug reports via jules chat or local HTTP feedback endpoints if available.
  • GitHub PR Integration: Leverages the official gh CLI to post rich, sanitized Markdown status boards directly on the active PR (gh pr comment <PR_ID> --body-file ...).
  • Graceful Fallbacks: If API tokens (GITHUB_TOKEN/GH_TOKEN) or the Jules environment are absent, the script outputs the exact payload to /tmp/jules_payload.md and displays it to stdout, completing successfully without failing the pipeline.

4. Human-in-the-Loop Developer Workflow

 +-----------------------------------------------------------------------------------+
 | 1. Developer asks Jules: "Generate and patch ASIMP OpenSCAP datastream retrieval" |
 +-----------------------------------------------------------------------------------+
                                          |
                                          v
 +-----------------------------------------------------------------------------------+
 | 2. Jules applies fixes, pushes branch, and generates GitHub Pull Request (e.g., #1) |
 +-----------------------------------------------------------------------------------+
                                          |
                                          v
 +-----------------------------------------------------------------------------------+
 | 3. Developer executes the local test suite on WSL2:                               |
 |    ansible-playbook -i inventory/hosts.yml playbooks/matrix_test.yml \            |
 |                     --extra-vars "execution_mode=dev pr_id=1"                     |
 +-----------------------------------------------------------------------------------+
                                          |
                                          v
 +-----------------------------------------------------------------------------------+
 | 4. Playbook orchestrates Podman containers, runs tests, and logs failing states   |
 +-----------------------------------------------------------------------------------+
                                          |
                                          v
 +-----------------------------------------------------------------------------------+
 | 5. Scripts automatically trigger, pushing telemetry to Jules & commenting on PR   |
 +-----------------------------------------------------------------------------------+
                                          |
                                          v
 +-----------------------------------------------------------------------------------+
 | 6. Developer reviews results in Jules and GitHub, giving next refactoring command |
 +-----------------------------------------------------------------------------------+

5. File Tree Structure

The local matrix integration introduces the following files to the codebase:

β”œβ”€β”€ ansible.cfg                          # Custom Ansible configuration with local roles path
β”œβ”€β”€ inventory/
β”‚   └── hosts.yml                        # Matrix hosts and execution variables
β”œβ”€β”€ playbooks/
β”‚   β”œβ”€β”€ matrix_test.yml                  # Playbook orchestrating Podman 5+ multi-OS containers
β”‚   └── roles/
β”‚       └── feedback_collector/
β”‚           └── tasks/
β”‚               └── main.yml             # Role assembling host metrics into structured JSON
└── scripts/
    └── jules_gh_feedback.sh             # Telemetry courier bridge script (Jules & GitHub PR)