Adlc plan

Skill davidvictor/adlc-skills/skills/adlc/adlc-plan

Agentic Development Lifecycle skills for real software work

Install

npx -y skills add davidvictor/adlc-skills --skill adlc-plan

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Create executable ADLC plans with task graphs, acceptance criteria, and commit checkpoints.

SKILL.md

1.9 KB, as published. Nobody here has run it

ADLC Plan

Use this to convert a clear request into an executable plan.

Modes

fast: one small task or quick fix; write .adlc/PLAN.md.
full: larger feature or single-sprint multi-step change; write .adlc/plans/<slug>.md.
parallel: only when tasks are independent enough for worktrees or Codex workers.
workstream: when an epic needs Codex goal-managed milestones, durable step cards, and multi-sprint state; use adlc-workstream.

Process

Resolve effective ADLC paths from .adlc/config.yaml; use adlc resolve-config when the CLI is available.
Read configured ADLC context, repo guidance, and any relevant spec under paths.interviews.
Run repo reconnaissance before asking questions.
If intent, non-goals, decision boundaries, or acceptance criteria remain materially unclear, hand off to adlc-interview before writing executable tasks.
Use plan-coordinator and plan-polisher when native Codex agents are available and the plan is non-trivial.
Choose the correct granularity: fast plan, full plan, parallel plan, roadmap item, or goal-managed workstream.
Define tasks with ownership, execution lane, dependencies, write scope, acceptance criteria, verification, docs/release impact, rollback notes when relevant, and commit checkpoints.
Split broad tasks until each task can be reviewed, verified, and committed as one coherent slice.
For 5 or more tasks, multiple milestones, or any long-running epic, create or update an adlc-workstream.
Mark blocked human decisions explicitly instead of hiding them inside implementation tasks.

Output

Write or return the plan path, interview spec path when used, task graph, granularity choice, verification policy, commit policy, open gates, and whether it is ready for adlc-improve, adlc-workstream, or adlc-implement.