Practical guides for teams building with coding agents.
PullPlane Resources is a field guide to shared specs, agent workflows, isolated execution, and human-reviewed software delivery.
Guides
Start with the operating principles behind agent-assisted software delivery.
AI Coding Agents for Software Teams
Coordinate Claude, Codex, Research, QA, and Security agents around a shared spec, isolated execution, and human-reviewed pull requests.
Read the guideAgentic Software Development with Human Review
Move from a collaborative spec through research, implementation, verification, and a human-approved pull request with multiple coding agents.
AI Project Management for Software Teams
Keep product decisions, engineering context, coding agents, branches, and pull-request review in one live software-delivery workflow.
AI Software Development Lifecycle Guide
A practical AI-assisted SDLC from team-authored spec through research, coding, testing, security review, and human-approved merge.
GitHub Project Management for Coding-Agent Teams
Connect live planning and agent execution to GitHub branches, checks, pull requests, reviews, and human-controlled merge state.
Coding agents
Understand the focused roles Claude, Codex, Research, QA, and Security can take on.
Claude Coding Agent for Team Workflows
Run Claude from a shared spec in an isolated sandbox, then review the diff, preview, and pull request with your team.
OpenAI Codex Coding Agent for Teams
Run OpenAI Codex against a shared card spec in an isolated sandbox and review its branch, diff, checks, and pull request.
Research Agent for Software Projects
Investigate a repository, technical risks, and implementation options with a read-only Research agent before code changes begin.
QA and Testing Agent for Regression Checks
Use a QA agent to test behavior, inspect edge cases, run useful checks, and report regressions against a shared acceptance spec.
Security Review Agent for Code Changes
Add a Security Reviewer agent to inspect trust boundaries, authorization, secrets, validation, injection risks, and unsafe data flows.
Workflows
Follow practical paths from a shared brief to reviewable, human-approved code.
Collaborative Spec to Pull Request Workflow
Co-edit a live engineering brief, run a coding agent in an isolated sandbox, and review the resulting branch and pull request.
AI-Assisted Bug Fix Workflow with Review
Capture a reliable reproduction, let a coding agent implement a focused fix, and verify the regression evidence before merge.
Research Before Code AI Agent Workflow
Use a read-only Research agent to map code paths, risks, and options before handing an agreed recommendation to a coding agent.
Multi-Agent Feature Development Workflow
Coordinate Research, Claude, Codex, and specialist review roles around one shared feature brief and one human merge decision.
Human-Gated AI Coding Agent Workflow
Let agents work in isolated sandboxes and open pull requests while preserving human approval and optional specialist checks at merge.
Security Review Agent Workflow for Pull Requests
Scope a security question, inspect the exact change for realistic risks, and keep remediation and risk acceptance in the human review flow.
QA Regression Testing Agent Workflow
Use a QA agent to trace acceptance criteria to checks, reproduce failures, inspect edge cases, and report test gaps before merge.
Technical Architecture Audit Agent Workflow
Bound an architecture question, gather repository evidence in read-only mode, compare options, and review the recommendation before implementation.
Recipes
Choose repeatable agent sequences for common engineering situations.
Research, Build, and Review AI Coding Recipe
Start ambiguous engineering work with read-only research, hand the recommendation to Codex, and finish with a structured review.
Focused Bug Fix and Review AI Recipe
Give Codex a reproducible bug and precise expected behavior, then send the smallest durable fix through a Reviewer pass.
UX Discovery, Build, and Review AI Recipe
Audit an existing interaction, let Claude implement the agreed UX change, and review accessibility and responsive behavior.
Technical Audit and Recommendation AI Recipe
Run a read-only architecture, performance, reliability, or security audit and challenge its evidence before code changes begin.
Integrations
See how repositories and isolated execution fit into the delivery loop.
GitHub Integration for AI Coding Agents
Connect selected GitHub repositories so PullPlane agents can work on card branches, open pull requests, and return to human review.
Forgejo Integration for AI Coding Agents
Connect a self-hosted Forgejo or compatible Gitea API instance so agents can work through repository branches and pull requests.
Daytona Sandboxes for AI Coding Agents
Run each PullPlane coding agent in a fresh Daytona sandbox with repository isolation, live logs, diffs, and preview handoff.