AID — AI-Integrated Development: A complete methodology for human-AI collaborative software development. 12 phases, 5 stages, 11 feedback loops.
This repository also configures its own agents. See what aid-methodology tells them →
Skill Claude Code
Revise or extend a document that already exists, in one pass. Use this skill when a document has drifted from the code, is missing a section, or needs a correction -- and you want the change drafted against the current source rather than written from memory. It reads the existing document first, then edits it…
Skill Claude Code
Update an existing infrastructure resource. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill plans and stops, so run /aid-execute to…
Skill Claude Code
Update an existing external-service integration. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill plans and stops, so run /aid-execute…
Skill Claude Code
Update an existing scheduled or background job. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill plans and stops, so run /aid-execute…
Skill Claude Code
Apply one targeted, human-confirmed change to the Knowledge Base. Use this skill when you know what changed and which part of the KB should reflect it, and you want that edit and nothing more. It works in its own worktree: it analyses how your instruction lands across the KB, turns that into a minimal scope plan with…
Skill Claude Code
Update an existing message/event schema or its emission. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill plans and stops, so run…
Skill Claude Code
Revise roadmap.md's ## MVP section only -- the first shippable slice: its contents, the line reasoning, what was cut, and its Status field (including the transition to Shipped ). Use this skill when the MVP section already carries committed content and its scope has changed. May create the ## MVP section if roadmap.md…
Skill Claude Code
Revise roadmap.md's direction entries outside the ## MVP section -- add, revise or supersede direction entries, and move an entry between horizon sections (## Now, ## Next, ## Later). Use this skill when the roadmap already exists and something has moved between horizons. Reads and consumes a roadmap seed when one is…
Skill Claude Code
Revise the project's technology (C0) Knowledge Base document -- languages, runtimes, frameworks, package managers, and build and test tooling with their versions, dependencies and constraints -- plus any previously created outputs you name. Use this skill when the technology-stack document already exists and a choice…
Skill Claude Code
Update or extend existing tests. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill plans and stops, so run /aid-execute to carry the…
Skill Claude Code
Revise the project's quality (C6) documents -- the test landscape (levels, coverage expectations, CI lane mapping, known gaps) and the gate policy (what blocks a merge, the thresholds, who may waive one) -- plus any previously created outputs you name. Use this skill when the testing-strategy record already exists and…
Skill Claude Code
Update an existing visual theme or style-token set. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill plans and stops, so run…
Skill Claude Code
Change exactly one part of an existing ticket in the project's issue tracker. Use this skill when a ticket needs its description replaced, a comment added, or its status moved: part is the closed enum description | comment | status, and one invocation touches one of them. It resolves which tracker owns the ticket…
Skill Claude Code
Update an existing UI component or page. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill plans and stops, so run /aid-execute to…
Skill Claude Code
Update an existing internal code artifact's behavior under new acceptance criteria. Use this skill when you already know what to update and want it scoped, specified, and broken into reviewable tasks in a single pass, with no requirements interview. You approve the resulting plan before anything is built: this skill…
Skill Claude Code
Regenerates the five install trees (claude-code, codex, cursor, copilot-cli, antigravity) from canonical/ + profiles/. Maintainer-only tooling; never shipped to end users. State machine: LOAD -> VALIDATE -> RENDER -> VERIFY -> REPORT.
Skill Claude Code needs its repo
Cut a release of THIS repo (the AID methodology repo) end-to-end: bump the version, update the release notes, gate, tag, publish to every channel, and verify all checks green. One level argument decides the bump: major | minor | patch | beta. Maintainer-only ops tooling — NOT an AID methodology skill, never rendered…
Skill Claude Code
Talk to another AI coding session. Use this skill when you want to ask a peer agent something, answer one that has asked you, or work alongside one -- whether it runs in this same tool or a different one. It gives you a channel: you open one, pull a named agent into it, send and read messages, and acknowledge what you…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: