Devflow

Open the program-aware TUI viewer — read-only terminal UI showing parallel sessions, org tree, and active initiatives in three stacked panels. tmux-safe; reflows narrow terminals. Triggers on: "open tui", "show tui", "what's running across sessions", "show org tree".

Orchestrate a multi-step DevFlow workflow by chaining skills. Use when the user wants to invoke a sequence of skills as one ask (e.g., "build and sync to github", "research, plan, then build", "ship and announce"). Triggers on: "ship X to Y", "build and X", "plan and X", "X then Y", "in one go", "as a chain", "all in sequence", "chain", "ship-and-sync", "research-plan-build"

Manage todos: capture an idea/task or view the morning standup across local + GitHub + peer sources. Subcommand-style: /devflow:todo add | list Use when the user wants to save something for later, note an idea, or get a "what should I work on?" view. Triggers on: "remember to", "add a todo", "save this idea", "what should I work on?", "morning standup", "check todos".

Investigate bugs using a structured approach that survives context resets. Tracks hypotheses, evidence, and progress across sessions. Use when the user reports a bug, error, or something not working as expected. Triggers on: "debug this", "something's broken", "fix this bug", "not working", "there's an error", "why isn't this working?"

Sub-30-LOC, single-file changes. The cheapest DevFlow path (~2k tokens). Use for typo fixes, single-line bug fixes, prop renames, dependency bumps, missing semicolons. Triggers on: "fix typo", "rename X to Y", "1-line fix", "single-file change", "tiny", "trivial"

Small features (single executor, no planner, no verifier) — between micro (1-line) and build (multi-subsystem). Cutoff: <5 files, <200 LOC, no new abstractions. Use when the change is too big for /devflow:micro but doesn't warrant full /devflow:build planning. Triggers on: "small change", "small feature", "5-file change", "isolated bug fix", "do this small task", "tackle this small change", "make a quick pass"

Resolve a parked decision and resume autonomous execution. Use when you see a pending DECISION-NNN.md in .planning/decisions/pending/, when the user wants to choose an option for a blocked checkpoint:decision, or when you need to tell the executor which option to take. Triggers on: "resolve decision", "decide DECISION-", "pick option for DECISION-", "unblock DECISION-", "answer DECISION-", "choose option for decision", "I want option-a", "go with option-b", "my answer is".

Consolidated status/resume/pause/health — `/devflow:status [resume|pause|check]`. Default: show progress + route to next action. Subcommands: check (integrity), pause (save context), resume (restore context). Both flag and bare forms accepted. Triggers on: "where are we?", "what's next?", "save my progress", "pick up where we left off".

Sync DevFlow planning state to GitHub — create/update objective issues, generate release notes, or push a single objective's state (body + sticky comment + Project v2 fields). Triggers on: "sync to github", "push objectives to github", "github release notes", "sync objective".

Hand off a TTY-required or shell-env-dependent command to the user's shell. Use for auth flows (gh auth login, doctl auth init, npm login) or shell-init commands (nvm/mise/conda activate, direnv). Non-disruptive when devflow-watch is running. Triggers on: "auth init", "log in to", "needs my password", "needs my shell environment".

Run the UI visual-evaluation pipeline on a Flutter surface or objective: capture each declared UI state and score it. Default (no --judge, or --judge labels) is an ADVISORY offline label-echo lookup (no network, no real vision comparison) — it can never clear a surface from human verification. Only --judge live (a real Anthropic vision call) produces a BINDING gate. Use when the user wants to visually evaluate, judge, or dogfood a UI surface's rendered states. Triggers on: "visual eval", "evaluate the UI", "ui-eval", "judge the screens", "check the visuals", "does this screen look right?"

Configure which agents run, which AI models they use, and how the workflow behaves. Use when the user wants to change DevFlow settings or configure workflow behavior. Triggers on: "change settings", "configure DevFlow", "update settings", "workflow settings"

Show cross-repo awareness: peer view (who's working on what in this repo) and org-wide progress (Product Roadmap). Renders both views by default. Triggers on: "who else is working on this", "what's in flight", "show org progress", "what are teammates working on".

Manage milestones: start a new one, audit a finished one, complete it, or plan gap-closure objectives. Subcommand-style: /devflow:milestone new | audit | complete | gaps Use when explicitly requested.

Add or remove an objective from the current milestone roadmap. Subcommand-style: /devflow:objective add | remove Use when explicitly requested. Note: 'insert' (decimal objectives) was removed in v1.2 — use 'add' instead.

Manage strategic initiative context — sync GitHub Epics to disk, list cached initiatives, or show a single initiative body. Planner reads these at plan time. Triggers on: "sync initiatives", "refresh initiatives", "show initiative", "what initiatives are loaded".

Set up a new project from scratch — asks what you're building, researches the domain, defines requirements, and creates a development roadmap. Use when the user wants to start a new project or initialize DevFlow for the first time. Triggers on: "start a new project", "initialize project", "let's build something new", "set up a new project", "new project"

Switch between quality, balanced, and budget model profiles to control cost and capability. Use when the user wants to change the AI model tier for DevFlow agents. Triggers on: "change profile", "switch to quality", "switch to balanced", "switch to budget", "set profile"

Run independent objectives in parallel using git worktrees, then merge results back. Subcommand-style: /devflow:workstreams setup | status | merge | run Advanced parallel execution — use only when explicitly requested.

Analyze an existing codebase to understand its stack, architecture, conventions, and concerns before starting new work. Use when the user wants to understand, analyze, or map an existing codebase. Triggers on: "understand this codebase", "map the code", "analyze architecture", "what does this codebase look like?", "explore the code structure"

Reconcile ROADMAP.md checkbox state against on-disk SUMMARY.md presence. Default: silently corrects drift; --dry-run previews; --interactive prompts per change. Triggers on: "sync roadmap", "reconcile roadmap", "fix roadmap drift", "is the roadmap accurate".

Run the ADVISORY UI design-critique sweep on a Flutter surface or objective: capture (or reuse) each declared UI state, critique it against the project design system through the design-critic engine, and present a prioritized design-debt list. This is advisory design critique — it NEVER gates. The heavy first-run produces a ranked design-debt backlog whose high-priority items feed the objective-work loop. Use when the user wants a design review, design critique, polish pass, or design-debt audit of a UI surface. Triggers on: "design review", "design-review", "critique the design", "design debt", "polish pass", "how's the design", "review the UI design"

Create a detailed execution plan for an objective, breaking it into concrete tasks with success criteria. Use when the user wants to plan an objective, create execution plans, or prepare for building. Triggers on: "plan objective", "create plans", "plan the next objective", "let's plan", "prepare objective", "make plans for"

Scan the codebase for security vulnerabilities — checks for secrets, auth flaws, dependency risks, and OWASP Top 10 issues. Standalone — works without project setup. Covers secrets, auth, dependencies, and code-level vulnerabilities. Triggers on: "security audit", "scan for vulnerabilities", "check for secrets", "security review", "find security issues"

Research how to build something before planning it — discovers best practices, architecture patterns, and pitfalls for the domain. Use when the user wants to investigate, research, or explore options before planning an objective. Triggers on: "research objective", "investigate before planning", "look into how to build", "what's the best approach for objective"

See what Claude plans to do for an objective before it starts, so you can course-correct early. Shows the intended approach without creating any files.