Govern coding-agent discovery and delivery through a project-local GAAI backlog
Drop a `.gaai/` folder into a project so agents separate scope discovery from isolated delivery, with acceptance criteria, memory, QA gates, and daemonized execution.
npx skills add agentskillexchange/skills --skill govern-coding-agent-discovery-and-delivery-through-a-project-local-gaai-backlog
.gaai/ folder into the target project or run the documented installer, then read .gaai/core/README.md, bootstrap the project, use /gaai-discover to create scoped stories, and use /gaai-deliver or /gaai-daemon for governed delivery.Use GAAI when a team wants coding agents to turn product intent into scoped stories first, then deliver only from a governed backlog with acceptance criteria and QA checks. The operator runs Discovery to clarify and write backlog stories, then runs Delivery or the daemon to execute stories in isolated agent sessions with planning, implementation, and QA roles. Invoke it instead of normal ad hoc coding-agent prompting when scope drift, forgotten decisions, or unverifiable acceptance criteria are recurring problems. The scope boundary is project-local governance for coding-agent discovery and delivery; it is not a general agent framework, SDK, or hosted workflow platform. Autonomous daemon delivery currently depends on Claude Code CLI, while discovery and governance are documented as portable across coding tools.
What this skill actually does
Inputs and prerequisites: AI coding tool, git, bash, python3; Claude Code CLI and tmux for autonomous delivery daemon.
Setup notes: Copy the repository’s .gaai/ folder into the target project or run the documented installer, then read .gaai/core/README.md , bootstrap the project, use /gaai-discover to create scoped stories, and use /gaai-deliver or /gaai-daemon for governed delivery.
Source and verification boundary: use https://github.com/Fr-e-d/GAAI-framework as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.