coding-backend-orchestration

Stars11

Forks3

UpdatedApril 11, 2026 at 14:37

Use when the user explicitly asks Guardian to launch or check an external coding assistant such as Codex, Claude Code, Gemini CLI, or Aider.

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Source

Threat-Vector-Security

Threat-Vector-Security/guardian-agent

View GitHub Repository View Creator Repositories

Download

Related occupationsSOC

Based on SOC occupation classification

Software DevelopersComputer and Mathematical Occupations·SOC 15-1252

File Explorer

2 files

SKILL.md

readonly

name	coding-backend-orchestration
description	Use when the user explicitly asks Guardian to launch or check an external coding assistant such as Codex, Claude Code, Gemini CLI, or Aider.

Coding Backend Orchestration

Overview / When to Use

Use this workflow only when the user explicitly wants Guardian to launch or inspect an external coding assistant.

This is distinct from normal repo-grounded coding work:

if the user asks Guardian to inspect, edit, patch, search, or test code directly, stay on the built-in coding tools
if the user explicitly says to use Codex, Claude Code, Gemini CLI, or Aider, launch that backend with coding_backend_run
mentioning a backend as the subject of a question is not enough to relaunch it

Process

Confirm the active coding session or list available sessions with code_session_current or code_session_list if the target workspace is unclear.
If needed, inspect available backends with coding_backend_list.
Launch the requested backend with coding_backend_run from the active coding session workspace.
If the user is asking whether a delegated run finished or what happened, use coding_backend_status.
After a backend run completes, verify the result with code_git_diff and the strongest relevant checks such as code_test, code_build, or code_lint.
Report backend output as delegated work, not as if Guardian performed the edits directly.

Red Flags

Treating "Why did Codex do that?" as permission to start a new Codex run.
Launching an external backend when the user only asked for direct repo inspection.
Reporting delegated backend work as successful without verification.
Running a backend outside the active coding session workspace.

Verification

The user explicitly requested an external coding backend.
The delegated run stayed anchored to the active coding session workspace.
You checked completion or recent runs with coding_backend_status when needed.
You verified the delegated result with code_git_diff, code_test, code_build, or code_lint before reporting success.

More from this repository

same repository

skill-creator

Threat-Vector-Security/guardian-agent

Guide for creating new skills, improving existing skills, and evaluating skill quality. Use this whenever the user wants to turn a workflow into a skill, revise a skill, improve triggering behavior, or add evals and benchmarks for a skill.

2026-04-1911

code-review

Threat-Vector-Security/guardian-agent

Use when reviewing a diff, PR, or patch for bugs, regressions, missing tests, architecture risk, security risk, or reviewability concerns before merge.

2026-04-1311

context-engineering

Threat-Vector-Security/guardian-agent

Use when starting a repo-grounded task, switching subsystems, or when output quality drops and the agent needs tighter project context from rules, specs, relevant files, examples, and current errors.

2026-04-1311

source-driven-development

Threat-Vector-Security/guardian-agent

Use when implementing or reviewing framework-specific code that should be verified against current official docs, versioned guidance, or source-cited examples rather than memory.

2026-04-1311

coding-workspace

Threat-Vector-Security/guardian-agent

Use when the request is about a repo, codebase, implementation, bugfix, or backend-owned coding session that should stay anchored to the active workspace.

2026-04-1111

receiving-code-review

Threat-Vector-Security/guardian-agent

Use when addressing code review or QA feedback, especially when the feedback is unclear, broad, or technically questionable.

2026-04-0411

name	coding-backend-orchestration
description	Use when the user explicitly asks Guardian to launch or check an external coding assistant such as Codex, Claude Code, Gemini CLI, or Aider.

Coding Backend Orchestration

Overview / When to Use

Use this workflow only when the user explicitly wants Guardian to launch or inspect an external coding assistant.

This is distinct from normal repo-grounded coding work:

if the user asks Guardian to inspect, edit, patch, search, or test code directly, stay on the built-in coding tools
if the user explicitly says to use Codex, Claude Code, Gemini CLI, or Aider, launch that backend with coding_backend_run
mentioning a backend as the subject of a question is not enough to relaunch it

Process

Confirm the active coding session or list available sessions with code_session_current or code_session_list if the target workspace is unclear.
If needed, inspect available backends with coding_backend_list.
Launch the requested backend with coding_backend_run from the active coding session workspace.
If the user is asking whether a delegated run finished or what happened, use coding_backend_status.
After a backend run completes, verify the result with code_git_diff and the strongest relevant checks such as code_test, code_build, or code_lint.
Report backend output as delegated work, not as if Guardian performed the edits directly.

Red Flags

Treating "Why did Codex do that?" as permission to start a new Codex run.
Launching an external backend when the user only asked for direct repo inspection.
Reporting delegated backend work as successful without verification.
Running a backend outside the active coding session workspace.

Verification

The user explicitly requested an external coding backend.
The delegated run stayed anchored to the active coding session workspace.
You checked completion or recent runs with coding_backend_status when needed.
You verified the delegated result with code_git_diff, code_test, code_build, or code_lint before reporting success.