All shelves / Mods / Workflow
ai-sdlc
By Sennjen
ai-sdlc is a Claude Code workflow mod by Sennjen.
Draws a band above the prompt with the feature's stage and one-click approvals, and denies code edits until you approve the intent, spec, and plan.
DrawsHolds a command
Install
The marketplace name is ai-sdlc.
claude plugin marketplace add Sennjen/claude-sdlc claude plugin install ai-sdlc@ai-sdlc
It runs inside the Claude Code process. The side panel is only what the author's README names. Read the repo before you install, or run claude plugin validate on the plugin folder.
- Version
- 0.12.0
- License
- MIT
- Repo stars
- 5
- Last change
- 2026-11-05
What it can reach
- Draws in the session
- Holds a command
- Sees tool calls
- Runs a command when you press
- Calls another program on this machine
Claude Code 2.1.293 or newer for the status bar (the gates work on older versions), python3 3.9 or newer, and git. After the install, run /ai-sdlc:init in your project, which detects the verify commands and writes sdlc.config.json, the switch that turns the plugin on for that repository; without it the plugin does nothing. Type sdlc approve intent, spec, or plan as a chat message to approve; only a message you type records an approval. In claude -p the module runs but draws nothing, so there is no band. In an interactive claude terminal the install is /plugin marketplace add Sennjen/claude-sdlc and /plugin install ai-sdlc@ai-sdlc.
The marketplace, ai-sdlc, holds only this plugin, and its renames map moves the earlier name sdlc to ai-sdlc. Approval gates from intent to review are why this page files it under workflow. interlock-renzrollon, for spec-driven development on OpenSpec, draws each wave and lane of a ship run, and refuses a test edit during the fix stage. learning-gate-2steaks refuses a code edit while a learning-guardrails checkpoint is open, and pins that question above the prompt. This one runs Intent, Spec, Plan, Build, Verify, and Review on files under docs/sdlc, and approval is a chat message you type, sdlc approve, recorded with the artifact's hash. Code edits wait for an approved intent, spec, and plan whose hashes still match, tests written for a bug fix stay read-only until you unlock them, and when a turn that changed code ends, your verify commands run and a failure sends Claude back. Independent design, code, and security reviewer agents read the work with no chat history. The band offers the one next action, such as Approve spec, Write plan, or Create PR.
Demo

What it hooks into
session.startWhen a session startsprompt.submitWhen you send a prompttool.callAround a tool call, where it can hold oneturn.completeWhen a turn endsui.renderDraws into the sessionSessionModeAbovePrompt
Read from the on() calls in its hook modules. Open its source under Files to check what each one does.
Hooks 9
UserPromptSubmitpython3PreToolUse · Edit|Write|MultiEdit|NotebookEditpython3PreToolUse · Bashpython3PostToolUse · Edit|Write|MultiEdit|NotebookEditpython3PostToolUse · Bashpython3PostToolUse · Bashpython3PostToolUse · ExitPlanMode|SubagentHandback|^mcp__.*(create|save|open)_?(pull|merge)_?requestpython3SubagentStop · sdlc-(security-|design-)?reviewer$python3Stoppython3
Files
The author's README and source, 4 files. They load when you open them.
Facts
- Author
- Sennjen
- Category
- Workflow
- Marketplace
- ai-sdlc
- Folder
- plugins/ai-sdlc
- Version
- 0.12.0
- License
- MIT
- Repo stars
- 5
- Last change
- 2026-11-05
- Language
- Python
- Files
- 112