Imported from sighthoundinc/sighthound-content-ops (deft/skills/deft-directive-setup/SKILL.md). Install upstream with npx skills add sighthoundinc/sighthound-content-ops --skill deft-directive-setup. Copyright stays with the author.

Deft Directive Setup

Agent-driven alternative to .deft/core/run bootstrap && .deft/core/run project && .deft/core/run spec.

Legend (from RFC2119): !=MUST, ~=SHOULD, ≉=SHOULD NOT, ⊗=MUST NOT, ?=MAY.

When to Use

User says "set up deft", "configure deft", or "bootstrap my project"
User asks to create USER.md, PROJECT-DEFINITION.vbrief.json, or a specification
User clones a deft-enabled repo for the first time with no config

Pre-Cutover Detection Guard

! Before proceeding with any setup phase, detect whether the project uses the pre-v0.20 document model and redirect to migration if so.

Detection Criteria

A project is pre-cutover if ANY of the following are true. This prose mirrors the executable helper in scripts/_precutover.py; when in doubt, the helper is canonical.

SPECIFICATION.md exists and is neither a deprecation redirect nor a current generated spec export. A current generated spec export contains  and , and vbrief/specification.vbrief.json plus all five lifecycle folders exist.
PROJECT.md exists and contains neither the legacy  sentinel NOR the current Purpose: deprecation redirect canonical-banner marker (same one-release-cycle grace window).
vbrief/specification.vbrief.json exists but the lifecycle folders (vbrief/proposed/, vbrief/pending/, vbrief/active/, vbrief/completed/, vbrief/cancelled/) do NOT exist

Action on Detection

! If pre-cutover state is detected, stop immediately and display an actionable message:

"This project uses the pre-v0.20 document model. Run task migrate:vbrief to upgrade to the vBRIEF-centric model."

! Include specific details about what was detected:

Missing lifecycle folders: "Run task migrate:vbrief to create the lifecycle folder structure"
SPECIFICATION.md with real content: "SPECIFICATION.md contains non-redirect content -- this file is deprecated; use scope vBRIEFs in vbrief/ instead"
PROJECT.md with real content: "PROJECT.md contains non-redirect content -- this file is deprecated; use PROJECT-DEFINITION.vbrief.json instead"
Missing PROJECT-DEFINITION.vbrief.json: "Run task project:render to generate the project definition"

Environment Preflight (before asking to run migration)

! Before asking the user "Would you like me to run task migrate:vbrief now?", run an environment preflight and report the results to the user. Do NOT ask the yes/no prompt until preflight results have been reported. Each failing check must be surfaced with a specific fix pointer so the user (or agent) can resolve the blocker before approving the run.

Run these four checks, in order:

Document-model confirmation -- re-apply the Detection Criteria above (the executable source is scripts/_precutover.py). If SPECIFICATION.md is a current generated spec export from vbrief/specification.vbrief.json and all lifecycle folders exist, stop: migration is NOT needed and MUST NOT be run.
Task resolvability -- check whether task migrate:vbrief is dispatchable from the project root:
- Run task --list (or platform-equivalent) and grep the output for a line containing migrate:vbrief.
- If present: the primary command works from the project root -- canonical invocation is task migrate:vbrief.
- If absent: the consumer Taskfile.yml does not include deft/Taskfile.yml. Fall back to the explicit-taskfile invocation task -t ./deft/Taskfile.yml migrate:vbrief and tell the user: "task migrate:vbrief is not resolvable from the project root. I will use the fallback invocation task -t ./deft/Taskfile.yml migrate:vbrief, which reads the task directly from the framework Taskfile. To make the primary command work in future, add an include for deft/Taskfile.yml to your project Taskfile.yml — see deft/main.md § Publishing deft tasks in your project root."
uv on PATH -- the migrator runs uv run python scripts/migrate_vbrief.py. Check uv --version (or equivalent): if it fails, point the user at the uv install docs (https://docs.astral.sh/uv/) and stop; migration cannot run without uv.
Migration script present -- check deft/scripts/migrate_vbrief.py exists on disk. If absent, the deft/ checkout is incomplete or came from a pre-v0.20 framework version; point the user at deft/QUICK-START.md (framework refresh guidance) and stop.

! Report each preflight check's result to the user (e.g. "✓ task migrate:vbrief resolvable", "✗ uv not on PATH — install from https://docs.astral.sh/uv/") BEFORE prompting for yes/no approval. If any check fails, do NOT offer to run migration until it is resolved.

⊗ Skip preflight and immediately ask "Would you like me to run task migrate:vbrief now?" -- preflight catches preventable errors (unresolvable task, missing uv, missing script) before the user commits to running migration. ⊗ Propose an install-step mutation that writes migrate:vbrief content into the consumer Taskfile. The supported publish mechanism is the includes: deft: deft/Taskfile.yml pattern documented in deft/main.md § Publishing deft tasks in your project root; inline Taskfile mutation is explicitly out of scope (per #506 D6).

Prompt and Run

! After preflight results are reported (and all checks pass), ask the user: "Would you like me to run task migrate:vbrief now?"

If yes: run the migration command (use the fallback invocation task -t ./deft/Taskfile.yml migrate:vbrief if the preflight resolvability check found the primary task unresolvable). Then re-run the pre-cutover detection guard to verify clean state before proceeding.
If no: stop and let the user handle migration manually.

⊗ Proceed with setup phases when pre-cutover artifacts are detected -- always redirect to migration first. ⊗ Silently ignore pre-cutover artifacts -- the user must be informed with an actionable command to fix the state. ⊗ Display the migration diagnostic without offering to run it -- always ask the user if they want the agent to handle it (after preflight has passed).

Greenfield Projects (No Migration Needed)

! For new projects (no existing SPECIFICATION.md, PROJECT.md, or vbrief/specification.vbrief.json), the guard passes silently and setup proceeds normally.

! Greenfield setup creates the full vBRIEF-centric structure from scratch:

./vbrief/ directory with all 5 lifecycle subdirectories: proposed/, pending/, active/, completed/, cancelled/
./vbrief/PROJECT-DEFINITION.vbrief.json generated from Phase 2 interview results
First scope vBRIEF created in proposed/ or pending/ depending on Phase 3 interview outcome

~ This is already handled by Phase 2 Output Path (creates ./vbrief/ and lifecycle subfolders) and Phase 3 Output (creates scope vBRIEFs in lifecycle folders). The guard ensures migrating projects are redirected before reaching these phases.

Migration safety flags

task migrate:vbrief is destructive by default (it replaces SPECIFICATION.md and PROJECT.md with redirect stubs and rewrites vbrief/), but it carries four always-on / on-demand safety affordances so operators can preview, recover, and undo (#497, #506 D7). Agents offering to run migration MUST mention these and pick the right one for the operator's situation.

! Automatic .premigrate.* backups (always-on, no flag): before any destructive write the migrator copies every pre-cutover input to its .premigrate sibling -- SPECIFICATION.md -> SPECIFICATION.premigrate.md, PROJECT.md -> PROJECT.premigrate.md, ROADMAP.md -> ROADMAP.premigrate.md, PRD.md -> PRD.premigrate.md (only if present), and vbrief/specification.vbrief.json -> vbrief/specification.premigrate.vbrief.json. Each backup emits a BACKUP <src> -> <dst> (<N> bytes) line in the migration output. These files are .gitignored by default so they do not leak into commits; operators who want them versioned can remove the patterns.
! task migrate:vbrief -- --dry-run (preview): prints the complete migration plan (every proposed backup, lifecycle folder, narrative ingestion, scope vBRIEF, and deprecation-redirect replacement) prefixed DRYRUN without writing any file. Exits 0 on success. Use this before running migration for the first time on an unfamiliar project.
! Dirty-tree guard (always-on): if git status --porcelain is non-empty the migrator refuses to run and points the operator at --force. Keeps migration output separable from in-progress edits. Bypass with task migrate:vbrief -- --force only after confirming the operator has accepted the risk.
! task migrate:vbrief -- --rollback: restores every pre-cutover input from its .premigrate.* backup and removes the scope vBRIEFs and migration-report files a prior run created. Reads vbrief/migration/safety-manifest.json (written by the migrator). Refuses if any redirect stub has been edited since migration -- re-run with --force to overwrite those edits on purpose, or commit them before rolling back. Prompts for a yes/no confirmation unless --force is passed.

! When a user declines an in-flight migration or reports an incorrect result, offer task migrate:vbrief -- --rollback before asking them to edit files by hand.

⊗ Offer task migrate:vbrief without also telling the user about --dry-run when they sound hesitant -- previewing is free and catches reconciliation surprises that would otherwise land in a commit. ⊗ Suggest git reset --hard or manual file deletion as a recovery path when --rollback would do the right thing more safely.

Deterministic Questions Contract

! Every numbered-menu prompt rendered in this skill (Phase 1 depth question, Phase 2 project type / deployment / language / strategy / branching gates, Phase 3 onboarding question, end-of-phase transition prompts, post-interview confirmation gate) MUST follow ../../contracts/deterministic-questions.md: render the canonical numbered menu in chat unless the host UI visibly preserves numeric option labels and returns numeric selections or exact displayed option text. The final two numbered options MUST be Discuss and Back, in that order. The Discuss-pause semantic is documented verbatim there -- on Discuss selection the agent MUST halt the in-progress sequence immediately, prompt What would you like to discuss?, and resume only on an explicit user signal (re-asks original question, says resume/continue, or re-issues prior selection). Implicit resumption is forbidden. Fallback chat replies MUST map only to the displayed number or exact displayed option text; do not infer from alphabetic host affordances.

Platform Detection

! Before resolving any config paths, detect the host OS from your environment context:

Platform	USER.md default path
Windows	`%APPDATA%\deft\USER.md` (e.g. `C:\Users\{user}\AppData\Roaming\deft\USER.md`)
Unix (macOS/Linux)	`~/.config/deft/USER.md`

! If $DEFT_USER_PATH is set, it takes precedence on any platform
! Create parent directories as needed when writing USER.md
~ $DEFT_PROJECT_PATH overrides the default project config path (./vbrief/PROJECT-DEFINITION.vbrief.json) if set

Agent Behavior

Flow:

! Start asking immediately — everything you need is in THIS file
⊗ Explore the codebase, read framework files, or gather context before asking
? Read deft/main.md or language files LATER when generating output

Interaction:

~ Use structured question tools only when their visible option labels preserve the canonical numbers (for example, 1. Yes) and their return value is the numeric selection or exact displayed option text.
! Fall back to the numbered chat menu when the host UI may replace numbers with alphabetic affordances, unlabeled buttons, or any other non-canonical choice labels.
⊗ Infer deterministic answers from host-added letters or shortcuts unless those letters were actually displayed in the canonical menu labels.

Defaults:

! Communicate that deft ships with best-in-class standards for 20+ languages
! Frame setup as "tell me your overrides" — not "configure everything"
~ "Deft has solid opinions on how code should be written and tested — I just need a few things about you and your project."

Adapt to Technical Level:

! First question gauges whether user is technical or non-technical
! Technical user: ask about languages, strategy, coverage directly — they'll have opinions
! Non-technical user: skip jargon, use sensible defaults, ask about what they're building not how
⊗ Ask non-technical users about coverage thresholds, strategies, or framework choices

Available Languages

C, C++, C#, Dart, Delphi, Elixir, Go, Java, JavaScript, Julia, Kotlin, Office.js (Excel JavaScript API), Python, R, Rust, SQL, Swift, TypeScript, VBA (Excel macros), VHDL, Visual Basic (.NET), Zig, 6502-DASM

? Read deft/languages/{name}.md when generating output — not before asking

Available Strategies

~ When presenting strategies to the user, always use this numbered list format (not a plain table). ~ Always include the chaining note below the list. ! Always show the FULL strategy list at every chaining gate — never remove a strategy because it was previously run. ~ If a strategy has been run already, indicate it with a note e.g. (run 1x) but keep it selectable.

interview ★ (recommended) — Structured interview with sizing gate: Light or Full path
yolo — Auto-pilot interview — Johnbot picks all recommended options
map — Analyze existing codebase conventions before adding features
discuss — Front-load decisions and alignment before planning
probe — Adversarially stress-test the plan; surface assumptions, edge cases, and risks before spec
research — Investigate the domain before planning
speckit — Five-phase spec-driven workflow for large/complex projects

💡 Strategies can be chained — after one completes, you'll be asked if you want to run another.

Phase 1 — User Preferences (USER.md)

Goal: Personal preferences file with two sections:

Personal — always wins over everything (name, custom rules)
Defaults — fallback values that PROJECT-DEFINITION.vbrief.json can override (strategy, coverage)
~ Skip if USER.md exists at the platform-appropriate path (see Platform Detection) and user doesn't want to overwrite
⊗ Scan filesystem beyond checking that one path

USER.md Freshness Detection

! When an existing USER.md is found (returning user), check its deft_version field before skipping Phase 1:

! If deft_version is missing: the USER.md predates versioning -- treat as stale
! If deft_version is present but differs from the current framework version (0.20.0): check whether any expected fields are missing from the USER.md
! If fields are missing: query the user for each missing field individually -- do NOT re-run the full Phase 1 interview
! After completing any field queries (even if none were needed), write the current deft_version (0.20.0) to USER.md
~ If deft_version matches the current version and all expected fields are present: no action needed (USER.md is fresh)

Expected USER.md fields: Name, Custom Rules, Default Strategy, and optionally Coverage and Experimental Rules.

⊗ Re-run the full Phase 1 interview when only individual fields are missing from a stale USER.md -- query missing fields individually instead.

Interview Rules

! This phase follows the deterministic interview loop defined in skills/deft-directive-interview/SKILL.md. The core rules (one question per turn, numbered options with stated default, explicit "other" escape, depth gate, default acceptance, confirmation gate, structured handoff) apply here. Key points repeated for emphasis:

! Each message you send MUST contain exactly ONE question. This is the most important rule in this file. After the user answers, send the NEXT question in a new message. Repeat until all questions for their track are answered.

⊗ Include two or more questions in the same message under any circumstances
⊗ List upcoming questions — only show the current one
~ Provide numbered answer options with an "other" choice where appropriate
! Mark which option is RECOMMENDED when showing choices
~ Use structured question tools only when visible option labels preserve the canonical numbers and returns map to numeric selections or exact displayed option text.

Question Sequence

Step 0 — Opening (all users): Ask: "How deep do you want to go?"

I'm technical — ask me everything
I have some opinions but keep it simple
Just pick good defaults — I care about the product, not the tools

Wait for answer. Then follow the track below.

Track 1 (technical) — 7 steps:

Step 1: Ask their name
Step 2: Ask strategy preference (show Available Strategies numbered list from the Available Strategies section, with descriptions and recommended marker; fallback — projects can override)
Step 3: Ask coverage threshold (default 85%; fallback — projects can override)
Step 4: Ask for custom rules — if user has rules, collect them one per line (empty line to finish); if none, skip
Step 5a: Present SOUL.md and ask whether to include it (default: yes):

SOUL.md — Results-first agent persona (inspired by Winston Wolf). Enforces assess-before-acting, finish-what-you-start, right-tool-for-the-job, and play-the-long-game. Keeps the AI decisive and concise. Includes a named persona ('Vinston') — drop if you prefer to define your own agent personality. Include SOUL.md? (Y/n)
Step 5b: Present morals.md and ask whether to include it (default: yes):

morals.md — Epistemic honesty rules. No presenting speculation as fact, label unverified claims, self-correct when wrong. Foundational trust rules for any AI agent. Strongly recommended. Include morals.md? (Y/n)
Step 5c: Present code-field.md and ask whether to include it (default: yes):

code-field.md — Pre-code assumption protocol. Requires stating assumptions and naming failure modes before writing a single line. Fights the 'it compiles, ship it' instinct. Based on NeoVertex1 context-field. Include code-field.md? (Y/n)

Track 2 (middle ground) — 2 steps:

Step 1: Ask their name
Step 2: Ask for custom rules — if user has rules, collect them one per line (empty line to finish); if none, skip
Set defaults without asking: strategy = "interview", coverage = 85%, all meta-guidelines included

Track 3 (non-technical) — 2 steps:

Step 1: Ask their name
Step 2: Ask what they're building (brief description — used for PROJECT-DEFINITION.vbrief.json later)
Set defaults: strategy = "interview", coverage = 85%, all meta-guidelines included

Output Path

Resolve using Platform Detection above. Write to the platform-appropriate path (or $DEFT_USER_PATH if set). Create parent directories as needed.

Template

# User Preferences

Legend (from RFC2119): !=MUST, ~=SHOULD, ≉=SHOULD NOT, ⊗=MUST NOT, ?=MAY.

**deft_version**: 0.20.0

## Personal (always wins)

Settings in this section have HIGHEST precedence — override all other deft rules,
including PROJECT-DEFINITION.vbrief.json.

**Name**: Address the user as: **{name}**

**Custom Rules**:
{custom rules or "No custom rules defined yet."}

## Defaults (fallback)

Settings in this section are fallback defaults. PROJECT-DEFINITION.vbrief.json overrides these
for project-scoped settings (strategy, coverage).

**Default Strategy**: [{strategy name}](../strategies/{strategy-file}.md)

{If coverage != 85: "**Coverage**: ! ≥{N}% test coverage"}

{If any experimental rules selected:
"## Experimental Rules

{one line per selected rule, e.g.:
- ! Use meta/SOUL.md for strategic context and purpose-driven guidance
- ! Use meta/morals.md for ethical AI development principles
- ~ Use meta/code-field.md for advanced architecture patterns}"}

---

**Note**: Edit this file anytime to update your preferences.
**See**: [../../main.md](../../main.md) for framework defaults.

Then

! Emit a structured-tool question asking whether to continue to Phase 2 (project configuration) only when the host preserves numeric labels; otherwise emit the deterministic numbered menu in chat. Options: 1. Yes (continue), 2. Not now (exit setup), 3. Discuss, 4. Back (revisit previous phase). The numeric labels MUST remain visible and be returned as numeric selections or exact displayed option text.
⊗ Ask the phase-transition question as unnumbered conversational prose or through a structured UI that hides the canonical numeric labels -- it is a deterministic menu and MUST preserve visible numbers (#478, #1563).

Phase 2 — Project Configuration (PROJECT-DEFINITION.vbrief.json)

Goal: Project-specific configuration — tech stack, type, quality standards — written as a vBRIEF file at ./vbrief/PROJECT-DEFINITION.vbrief.json.

! Path Resolution Anchor: Resolve ALL paths relative to the user's working directory (pwd) at skill entry -- never relative to the skill file location, AGENTS.md location, or any framework directory (e.g. ./deft/). When deft is cloned as a subdirectory, the skill file lives inside the clone but all project artifacts (./vbrief/PROJECT-DEFINITION.vbrief.json, build files, etc.) must be resolved from the user's pwd.

~ Skip if ./vbrief/PROJECT-DEFINITION.vbrief.json exists (or $DEFT_PROJECT_PATH if set) and user doesn't want to replace
⊗ Count ./deft/PROJECT-DEFINITION.vbrief.json or ./deft/core/project.md as the user's project config — those are framework-internal

Inference

! Before asking, infer from codebase — look for package.json, go.mod, requirements.txt, Cargo.toml, pyproject.toml, *.csproj
! Use inferences to pre-fill answers and confirm — don't ask blind
⊗ Look inside ./deft/ for build files (go.mod, package.json, pyproject.toml, Cargo.toml, *.csproj, etc.) — those are framework-internal. Only inspect files at the project root and its non-deft subdirectories.
⊗ Run git commands inside ./deft/ to determine project identity — that directory is the framework repo, not the user's project.
~ If no build files are found at the project root, default the project name to the current directory name and ask for confirmation.

Track Detection

! If Phase 1 was skipped (USER.md already existed), the user's track is unknown. Before asking any Phase 2 questions, ask the depth question:

"How deep do you want to go?"

I'm technical — ask me everything

I have some opinions but keep it simple

Just pick good defaults — I care about the product, not the tools

Wait for answer. Then follow the corresponding track in the Question Sequence below.

⊗ Assume Track 1 (technical) because USER.md exists or contains strategy/coverage fields. ⊗ Infer the track from USER.md content — always ask.

Defaults in Agentic Mode

! When a question has a USER.md default, phrase it as:

"{Field}: {value} from USER.md — keep this, or enter a different value?"

! Accept any affirmative response ("keep", "yes", "same", "default", ✓) as confirmation to use the default. ⊗ Phrase defaults as "press Enter to keep" — there is no Enter in conversational mode.

Interview Rules (same as Phase 1)

! Each message MUST contain exactly ONE question. The Phase 1 interview rules apply here too. Do not combine questions. See skills/deft-directive-interview/SKILL.md for the canonical deterministic interview loop.

Question Sequence

Track 1 (technical) — 8 steps:

Step 1: Ask project name (infer from build files or directory name, confirm)
Step 2: Ask project type (CLI, TUI, Desktop App, REST API, Web App, Library, other)
Step 3: Ask deployment platform:
1. Cross-platform (Linux / macOS / Windows)
2. Windows-native
3. macOS-native
4. Linux / Unix
5. Embedded / low-resource
6. Web / Cloud
7. Mobile (iOS / Android)
8. Other / not sure
Step 4: Ask languages — show a filtered shortlist (3–4 recommendations) based on project type + platform. If codebase markers exist (go.mod, pyproject.toml, etc.), skip and confirm: "Detected {lang} — correct?"
- If user selects "Other": show remaining plausible languages for the type+platform context (Tier 2)
- If still not found: free text input (Tier 3)
- If entered language has no deft languages/{lang}.md standards file, warn: "deft doesn't have a standards file for {lang} yet — general defaults will be used. Continue?"
Step 5: Ask tech stack (frameworks, libraries)
Step 6: Ask strategy (default to USER.md Defaults; ask if this project needs different — show Available Strategies numbered list with descriptions and recommended marker)
Step 7: Ask coverage (default to USER.md Defaults; ask if this project needs different)
Step 8: Ask for project-specific rules (optional, same one-per-line format as Phase 1 custom rules)
Step 9: Ask branching preference (typed plan.policy.allowDirectCommitsToMaster flag per #746):

! Render this as a deterministic numbered menu. Default 1. Branch-based. Final two options MUST be Discuss and Back per ../../contracts/deterministic-questions.md:
"Do you prefer branch-based workflow (create a feature branch for every change) or trunk-based (commit directly to master)? Branch-based is the default and recommended for teams; trunk-based is common for solo projects."
1. Branch-based ★ (recommended -- default; enforces feature branches via the deft branch-protection policy)
2. Trunk-based (direct commits to master) -- see capability-cost disclosure below
3. Discuss
4. Back
! Capability-cost disclosure (#746): When the user picks option 2 (trunk-based), the agent MUST present the capability-cost disclosure verbatim BEFORE writing the typed flag, then re-prompt for explicit confirmation:
"Capability-cost disclosure -- enabling direct commits to the default branch turns OFF the deft branch-protection policy. The pre-commit + pre-push hooks will no longer block default-branch commits, task verify:branch will pass on the default branch, and the skill-level guards in deft-directive-{swarm,review-cycle,pre-pr,release} will not halt for default-branch work. The change is reversible (task policy:enforce-branches) and is recorded to meta/policy-changes.log for auditability. The CI sanity check (head_ref != base_ref) remains independent and will continue to flag master->master PRs. Are you sure?"
1. Yes, opt out -- write plan.policy.allowDirectCommitsToMaster = true
2. No, keep branch-protection enforced -- write plan.policy.allowDirectCommitsToMaster = false
3. Discuss
4. Back
! Default to option 2 (enforce). Explicit affirmative on option 1 is required to opt out -- a broad proceed does NOT satisfy this gate. The same affirmative-only rule applies as in /deft:change (yes, confirmed, approve).

! Write the answer to plan.policy.allowDirectCommitsToMaster (typed boolean) on the PROJECT-DEFINITION vBRIEF. Default false (enforce branches) when the user picks option 2 OR omits the question entirely. Writing this typed surface is what the framework reads going forward; agents MUST NOT write the legacy free-form Allow direct commits to master: narrative key (#746 part A migrates the legacy narrative away).

! Re-running the interview detects the existing flag (#746 part G2): If vbrief/PROJECT-DEFINITION.vbrief.json already exists and has plan.policy.allowDirectCommitsToMaster set, the interview MUST surface the current value (e.g. "Current setting: allowDirectCommitsToMaster=false (branch-protection ON)") and ask whether to keep it or change it before re-prompting. Do not silently overwrite an existing typed value.

! Slash-command alternatives (#746 part G2): Once the project is set up, the typed flag can also be flipped via slash commands wrapping task policy:*:
- /deft:policy:show -- display the current resolved policy and source
- /deft:policy:enforce-branches -- set allowDirectCommitsToMaster=false
- /deft:policy:allow-direct-commits -- set allowDirectCommitsToMaster=true (requires --confirm to apply)
Each transition is recorded to meta/policy-changes.log for auditability.

Track 2 (middle ground) — 4 steps:

Step 1: Ask project name (infer from build files or directory name, confirm)
Step 2: Ask project type (CLI, TUI, Desktop App, REST API, Web App, Library, other)
Step 3: Ask languages (show detected, confirm or adjust; if none detected, infer from type and ask)
Step 4: Ask strategy (default to USER.md Defaults; ask if this project needs different — show Available Strategies numbered list with descriptions and recommended marker)
Default coverage to USER.md Defaults without asking

Track 3 (non-technical) — 1 step:

Step 1: Present summary of inferences: "Based on your project: {name} ({type}), built with {stack}. Look right?"
⊗ Ask about strategy or coverage — use Phase 1 defaults

Output Path

./vbrief/PROJECT-DEFINITION.vbrief.json (or $DEFT_PROJECT_PATH if set). Create ./vbrief/ directory and lifecycle subfolders (proposed/, pending/, active/, completed/, cancelled/) if they don't exist.

GitHub PR Template Scaffolding (#531)

! Before writing PROJECT-DEFINITION.vbrief.json, offer to scaffold a default GitHub PR template so downstream skills (deft-directive-refinement Pre-Flight, deft-directive-pre-pr) can satisfy their .github/PULL_REQUEST_TEMPLATE.md checks without blocking.

! Ask the user with a deterministic numbered menu: "Create a default GitHub PR template at .github/PULL_REQUEST_TEMPLATE.md?" Options: 1. Yes, 2. No, 3. Discuss, 4. Back. Use a structured question tool only if those numeric labels remain visible and are returned as numeric selections or exact displayed option text.
! If the user accepts AND .github/PULL_REQUEST_TEMPLATE.md does NOT already exist: copy templates/PULL_REQUEST_TEMPLATE.md (shipped with deft) to ./.github/PULL_REQUEST_TEMPLATE.md in the consumer project. Create .github/ if it does not exist.
! If the file already exists, do NOT overwrite it — report that it is present and continue.
~ If the user declines, note that deft-directive-refinement Pre-Flight will offer to scaffold later when needed.

⊗ Overwrite an existing .github/PULL_REQUEST_TEMPLATE.md without explicit user approval.

Headless Coverage Warning — display-bound GUI entry points (#1027)

! The trigger is a display-bound GUI event loop (pygame, tkinter, PyQt/PySide, Kivy, Electron) that cannot run without a real display — typically a Desktop App project type, or a TUI that embeds such a GUI. Terminal-UI frameworks (textual, urwid, blessed, ncurses) run in the terminal and DO support headless testing (e.g. textual's App.run_async() + Pilot), so a standard TUI is NOT in scope — do not omit its coverage. The concrete commands below assume a Python GUI stack (pygame/tkinter); the same "omit the un-runnable loop, test the logic" principle applies to non-Python desktop stacks (Electron/JS, .NET/WPF, Qt/C++) using that language's own headless-test and coverage-exclusion tooling. When the Phase 2 project type resolves to a display-bound GUI project, warn the user BEFORE writing PROJECT-DEFINITION.vbrief.json (adapt the wording to the project's language):

"Heads up: pygame/tkinter event loops can't be tested headlessly, so the display-bound entry point (e.g. src/ui.py) reports near-zero coverage and drags the overall percentage below the 85% threshold. I recommend excluding the UI entry point from coverage measurement and keeping it thin — push testable logic (state, scoring, input handling) into separate modules."

! When scaffolding or advising on pyproject.toml for a display-bound GUI project, add the display-bound entry point to [tool.coverage.run] omit so task check measures logic modules only:

[tool.coverage.run]
omit = [
  "*/tests/*",
  "*/venv/*",
  "*/.venv/*",
  "src/ui.py",        # display-bound pygame/tkinter event loop -- cannot run headlessly (#1027)
]

! Keep the omit narrow — exclude only the event-loop shell, never a module that also holds business logic. If logic and the loop are mixed, recommend refactoring the logic into a separate, fully-tested module first.
~ For a Python project, point the user at languages/python.md (the Headless GUI / event-loop testing section under Patterns) for the headless-test pattern (SDL_VIDEODRIVER=dummy) and the full coverage-omit rationale; for a non-Python GUI stack, apply the same principle with that language's headless-test and coverage-exclusion tooling.
⊗ Apply the omit to a headless-capable terminal-UI project (textual/urwid/blessed/ncurses) — those frameworks test headlessly, so omitting them hides measurable coverage, the opposite of the intended effect.
⊗ Silently accept the default 85% coverage gate for a display-bound GUI project without surfacing the headless blind spot — the agent reports an inflated per-session coverage that collapses when the full src/ is measured (the 2026-05-10 tic-tac-toe desktop-UI swarm recurrence).

Template

! The output MUST conform to the canonical vBRIEF v0.6 schema (vbrief/schemas/vbrief-core.schema.json, strict const: "0.6"). See ../../conventions/references.md.

{
  "vBRIEFInfo": {
    "version": "0.6",
    "author": "agent:deft-directive-setup",
    "description": "Project identity gestalt",
    "created": "{ISO-8601 timestamp}"
  },
  "plan": {
    "title": "{Project Name}",
    "status": "running",
    "narratives": {
      "Overview": "{Brief project description}",
      "TechStack": "{project type} using {languages} — {tech stack details}",
      "Strategy": "Use {strategy name} for this project",
      "Quality": "Run task check before every commit. Achieve >= {coverage}% coverage overall + per-module. Store secrets in secrets/ dir.",
      "ProjectRules": "{Any rules the user specified, or 'No project-specific rules defined.'}",
      "Branching": "{If trunk-based: 'Allow direct commits to master: true', else omit or 'Branch-based workflow (default)'}",
      "DeftVersion": "0.20.0"
    },
    "items": []
  }
}

! All narratives values MUST be plain strings — never objects or arrays
! items starts empty — populated as scope vBRIEFs are created in lifecycle folders

Then

! Emit a structured-tool question asking whether to continue to Phase 3 (specification) only when the host preserves numeric labels; otherwise emit the deterministic numbered menu in chat. Options: 1. Yes (continue), 2. Not now (exit setup), 3. Discuss, 4. Back (revisit previous phase). The numeric labels MUST remain visible and be returned as numeric selections or exact displayed option text.
⊗ Ask the phase-transition question as unnumbered conversational prose or through a structured UI that hides the canonical numeric labels -- it is a deterministic menu and MUST preserve visible numbers (#478, #1563).

Follow-up: triage onboarding (#1143)

~ After Phase 2 writes PROJECT-DEFINITION.vbrief.json, recommend task triage:welcome to the user as the single chained command for picking up the v0.27 triage surface. The N3 ritual (#1143) is the consolidating onboarding step for the #1119 governance swarm verbs (task triage:bootstrap / task triage:scope / plan.policy.wipCap writes / task scope:demote --batch relief / task triage:summary); without it consumers must learn each verb individually from the v0.27 release notes.
~ task triage:welcome is idempotent and detection-bound -- each phase emits an informational stderr line and skips when its precondition is already satisfied, so a re-run after a partial completion resumes cleanly. The destructive phases (subscription / wipCap writes, optional WIP-relief invocation) are gated by numbered-menu prompts per ../../contracts/deterministic-questions.md. See ../../UPGRADING.md ## From v0.26.x -> v0.27 for the full walkthrough.
? The recommendation is informational, not a hard gate -- consumers who plan to wire triage manually MAY skip the ritual and call the individual verbs in any order; the framework defaults stay fail-open per the umbrella #1119 §12 framework-vs-consumer-config boundary.

Phase 3 — Specification

Goal: Generate an implementable spec using the strategy chosen in Phase 2, producing a specification.vbrief.json draft for human approval before downstream generation.

! Path Resolution Anchor: Same rule as Phase 2 -- resolve ALL paths relative to the user's pwd at skill entry, never relative to the skill file, AGENTS.md, or any framework directory.

~ Skip if user already has scope vBRIEFs in ./vbrief/ they're happy with
! Check ./vbrief/specification.vbrief.json or ./vbrief/proposed/ for existing scope vBRIEFs
⊗ Count ANY file inside ./deft/ as the project's spec — those are framework-internal (e.g. deft/PROJECT.md, deft/specs/, deft/templates/, deft/core/project.md are all part of the framework, NOT the user's project)

Onboarding Question

! Before proceeding with the strategy gate, ask the onboarding question:

"Are you adding a scope to this project or starting a new specification?"

Adding scope to existing project [default if ./vbrief/specification.vbrief.json exists or scope vBRIEFs found in lifecycle folders]

Starting a new project specification [default if no specification or scope vBRIEFs exist]

! Default based on repo state: if specification.vbrief.json exists or any lifecycle folder has scope vBRIEFs, default to "Adding scope"; otherwise default to "Starting new"
! If adding scope: skip the full interview, create a new scope vBRIEF in ./vbrief/proposed/ with the user's description, then exit
! If starting new: proceed to the Strategy Gate below

⚠️ MANDATORY: Strategy Gate — Do This First

! STOP. You MUST determine the correct strategy before doing anything else.

! Open ./vbrief/PROJECT-DEFINITION.vbrief.json (the file written in Phase 2)
! Find the narratives.Strategy value
! Extract the strategy name from the narrative

Dispatch:

interview (or default) → Continue to the Sizing Gate below ✅
anything else (discuss, yolo, speckit, research, brownfield, map, etc.) →
1. ! Read deft/strategies/{strategy-name}.md right now, in this same turn
2. ! Begin the strategy's workflow immediately — ask its first question
3. ! STOP reading this section — do NOT use the interview process below
⊗ Default to interview without reading PROJECT-DEFINITION.vbrief.json
⊗ Continue reading below when PROJECT-DEFINITION.vbrief.json specifies a non-interview strategy
⊗ Assume interview because the sections below describe the interview process
⊗ Fabricate justification for using interview when the user chose a different strategy
⊗ Announce the strategy choice and then stop — you must immediately read the file and start

⬇️ Everything below applies ONLY to the interview strategy. If your strategy is anything else, STOP — follow your strategy file instead.

Sizing Gate (interview and yolo strategies only)

! After hearing what the user wants to build and their feature list, determine project complexity per strategies/interview.md.

! Check PROJECT-DEFINITION.vbrief.json narratives for Light or Full — if declared, use that path
! If not declared, propose a size and ask the user to confirm in a dedicated message
! Wait for the user's response before asking any interview questions
⊗ Combine the sizing proposal with the first interview question
⊗ Proceed to interview questions before the user has confirmed the path

Light (small/medium): Interview → specification.vbrief.json with slim narratives (Overview + Architecture) → scope vBRIEFs in vbrief/proposed/. Full (large/complex): Interview → rich narratives in specification.vbrief.json (user approval) → scope vBRIEFs with traceability.

Interview Process (interview strategy)

Per strategies/interview.md:

! Ask what to build and features first
! Ask ONE focused, non-trivial question per step
~ Provide numbered options with an "other" choice
! Mark which option is RECOMMENDED
⊗ Ask multiple questions at once
⊗ Make assumptions without clarifying
~ Use structured question tools for interview questions only when they preserve visible numeric option labels and return numeric selections or exact displayed option text; otherwise render the numbered menu in chat.

Question Areas:

! Missing decisions (language, framework, deployment)
! Edge cases (errors, boundaries, failure modes)
! Implementation details (architecture, patterns, libraries)
! Requirements (performance, security, scalability)
! UX/constraints (users, timeline, compatibility)
! Tradeoffs (simplicity vs features, speed vs safety)

Non-Technical Users:

~ Adjust vocabulary: "How do you want to store data?" not "What database engine?"
~ "Will other apps talk to this?" not "REST or GraphQL?"

Completion:

! Continue until little ambiguity remains
! Spec must be comprehensive enough to implement

Output — Light Path

! Write ./vbrief/specification.vbrief.json with "vBRIEFInfo": { "version": "0.6" }, status: draft, and slim narratives:
- Overview: Brief project summary
- Architecture: System design description
! Create scope vBRIEFs in ./vbrief/proposed/ for each identified work item
- Each scope vBRIEF follows the YYYY-MM-DD-descriptive-slug.vbrief.json filename convention (slug rules in ../../conventions/vbrief-filenames.md)
- Each MUST use "vBRIEFInfo": { "version": "0.6" }
- Each MUST include embedded Requirements (FR-N, NFR-N) in its narrative
- Each task SHOULD reference which FR/NFR it implements via narrative.Traces
- When the scope originates from a GitHub issue, include a references entry in the canonical form (see ../../conventions/references.md):
```
"references": [
  {
    "uri": "https://github.com/{owner}/{repo}/issues/{N}",
    "type": "x-vbrief/github-issue",
    "title": "Issue #{N}: {issue title}"
  }
]
```
! Summarize decisions, ask user to review the vBRIEF narratives
! On approval, update specification.vbrief.json status to approved

⊗ Create a separate PRD.md on the Light path
⊗ Generate an authoritative PRD.md — if needed, users run task prd:render

! The vBRIEF files MUST conform to vbrief/schemas/vbrief-core.schema.json (v0.6):

! All narratives and narrative values MUST be plain strings — never objects or arrays
! Nested children within a PlanItem use items (v0.6 preferred field); subItems is the deprecated legacy alias kept for backward compatibility only
⊗ Mix items and subItems on the same PlanItem — pick one (prefer items)

Output — Full Path

! Write rich narratives to ./vbrief/specification.vbrief.json with "vBRIEFInfo": { "version": "0.6" }, plan.status: draft, and these narrative keys:
- ProblemStatement: What problem this project solves
- Goals: High-level project goals
- UserStories: User stories in standard format
- Requirements: Structured requirements (FR-N: ..., NFR-N: ...)
- SuccessMetrics: Measurable success criteria
- Architecture: System design and technical architecture
- Overview: Brief project summary
! Human approval gate: Present the vBRIEF draft narratives to the user for review — reviewing the specification.vbrief.json narratives IS the approval step (replaces the former PRD.md review). The user may request changes before approving.
! On approval, update status to approved and proceed to downstream generation
! Create scope vBRIEFs in ./vbrief/proposed/ with traceability to requirement IDs from the narratives

! Scope vBRIEFs MUST trace tasks back to requirement IDs (FR-1, NFR-1) from the Requirements narrative
⊗ Generate an authoritative PRD.md — if needed, users run task prd:render

Spec Structure (both paths):

! Overview, Architecture
! Implementation Plan: scope vBRIEFs in vbrief/proposed/ with phases and dependencies
! Explicit dependency mapping between scopes (via vBRIEF edges or references)
~ Scopes designed for parallel work by multiple agents
! Testing Strategy and Deployment captured in narratives
⊗ Write code — specification only

Lifecycle Bridge to Downstream Skills (#1025)

! Scope vBRIEFs created by Phase 3 (both Light and Full paths) AND by the Onboarding Question "Adding scope to existing project" branch land in vbrief/proposed/ with plan.status: proposed. This is the canonical deposit point per the deft lifecycle (proposed -> pending -> active -> completed). The #810 implementation-intent gate (task vbrief:preflight) and the deft-directive-swarm Phase 0 Step 1 preflight BOTH require candidate vBRIEFs to live in vbrief/active/ with plan.status == "running" before any agent can dispatch against them; setup deliberately stops at proposed/ because the lifecycle commitment (promote + activate) belongs to the downstream skill, not the setup interview.

! Surface this bridge to the user in the Phase 3 → next-skill handoff so they are not surprised by a wholesale preflight rejection downstream:

If the next step is skills/deft-directive-swarm/SKILL.md: the swarm skill's Phase 0 Step 0.5 (Lifecycle Bridge -- Promote and Activate Proposed Scope vBRIEFs) is the canonical bridge. The monitor will scan vbrief/proposed/ and vbrief/pending/, present in-scope candidates, and run task scope:promote -- <path> then task scope:activate -- <path> on explicit user approval. No manual operator action is required ahead of the swarm invocation.
If the next step is skills/deft-directive-refinement/SKILL.md: the refinement skill's Phase 4 (Promote/Demote) owns the same task scope:promote / task scope:activate surface and runs the bridge as part of the refinement loop. The refinement skill MAY leave vBRIEFs in pending/ deliberately when they are queued for prioritisation rather than immediate dispatch.
If the user wants to invoke an implementation agent directly via skills/deft-directive-build/SKILL.md or start_agent: the bridge MUST be run manually before dispatch -- task scope:promote -- vbrief/proposed/<file> then task scope:activate -- vbrief/pending/<file>. Both commands are idempotent and exit 0 on no-op (see scripts/scope_lifecycle.py). The #810 preflight gate (task vbrief:preflight -- <active-path>) will exit 0 only after the activate step.

⊗ Auto-run task scope:promote or task scope:activate from the setup skill on the Phase 3 outputs. The lifecycle commitment belongs to the user ("I am ready to swarm/build on this scope"), not the setup interview; silent promotion would clear the #810 implementation-intent gate without explicit user authorisation and bypass the deterministic-questions contract that protects every other Phase 3 transition.

⊗ Drop the user at the end of Phase 3 with scope vBRIEFs in vbrief/proposed/ and no forward pointer to the bridge. Without this section the user discovers the gap at runtime when the swarm Phase 0 Step 1 preflight rejects every candidate (Invalid transition: 'activate' requires file in pending/), as in the originating 2026-05-10 first-session consumer tic-tac-toe swarm (issue #1025).

End-of-Phase-3 Export Prompt and Render Gate

! After the human approval gate on specification.vbrief.json narratives but BEFORE handing off to deft-directive-build (or advancing speckit Phase 3 → Phase 4), ask the user whether to generate human-readable exports. This replaces the invisible skip-if-absent behavior of task check (#398) and closes the greenfield gap (#433). This is also the Phase 3 → Phase 4 transition gate required by strategies/speckit.md Post-Phase 3 Transition Gate (#432).

! Prompt: "Your specification.vbrief.json is approved. Generate SPECIFICATION.md and/or PRD.md now? (recommended for stakeholder review)"
1. Yes — render both
2. SPECIFICATION.md only
3. PRD.md only
4. Skip — I’ll render later with task spec:render / task prd:render
! Run the selected render command(s):
- task spec:render → writes SPECIFICATION.md
- task prd:render → writes PRD.md
! If the user picked a speckit-strategy project: task spec:render is mandatory at this boundary — invoke it even if the user declined the prompt, because speckit Phase 3 → Phase 4 is gated on SPECIFICATION.md existing and matching the current vBRIEF hash.
! Confirm to the user which files were written and remind them that direct edits to SPECIFICATION.md / PRD.md are overwritten on the next render — edit specification.vbrief.json instead.
~ If the user skipped rendering and is NOT on a speckit strategy, no-op and continue.

⊗ Advance a speckit project to Phase 4 without running task spec:render at this gate — SPECIFICATION.md is required for the Phase 3 transition criterion. ⊗ Silently skip the prompt — greenfield users who never open a PR will miss the exports without it.

Handoff to deft-directive-build

! Emit a structured-tool question asking whether to continue to the build phase only when the host preserves numeric labels; otherwise emit the deterministic numbered menu in chat. Options: 1. Yes (continue), 2. Not now (exit setup), 3. Discuss, 4. Back (revisit previous phase). The numeric labels MUST remain visible and be returned as numeric selections or exact displayed option text.
~ If platform supports skill invocation and the user picks Yes, invoke skills/deft-directive-build/SKILL.md
⊗ Leave user with a dead end -- always offer the next step via the structured-tool phase-transition question
⊗ Ask the handoff-to-build question as unnumbered conversational prose or through a structured UI that hides the canonical numeric labels -- it is a deterministic menu and MUST preserve visible numbers (#478, #1563).

Warp Auto-Approve Warning

! Recommended Warp setting: Before running deft-directive-setup, ensure Warp's AI autonomy is set to "Always ask" in AI -> Profile Settings. When set to a higher autonomy level (e.g. "Auto-run"), Warp may silently self-answer interview questions without user input, producing garbage USER.md/PROJECT-DEFINITION.vbrief.json with no error or warning. The post-interview confirmation gate (below) is the last line of defense, but prevention is better than detection.

Post-Interview Confirmation Gate

! After completing ALL interview questions for any phase (Phase 1, Phase 2, or Phase 3), but BEFORE writing any files:

! Display a summary of all captured values in a clearly formatted list -- include every field that will be written to the output file (e.g. name, strategy, coverage, languages, project type, custom rules, etc.)
! Ask the user for explicit confirmation: "These are the values I captured. Write files? (yes/no)"
! Accept only explicit affirmative responses (yes, confirmed, approve) -- reject vague responses (proceed, do it, go ahead) the same way /deft:change does
! If the user says no: re-display the values and ask which ones to correct, then re-confirm before writing
! If any value appears to be auto-generated filler (e.g. repeated default text, placeholder strings, or values that echo the question prompt), warn the user explicitly: "Some values look like they may have been auto-filled rather than provided by you. Please review carefully."

⊗ Write USER.md, PROJECT-DEFINITION.vbrief.json, specification.vbrief.json, or any other deft-directive-setup artifact without first displaying captured values and receiving explicit user confirmation. ⊗ Treat a broad "proceed" or "continue" as confirmation to write files -- the user must explicitly confirm the displayed values.

? Yolo strategy carve-out: When the user's chosen strategy is yolo (auto-pilot), the confirmation gate still applies but the agent (Johnbot) may self-confirm on the user's behalf by displaying the summary and immediately proceeding -- the user has already opted into auto-pilot by selecting yolo. The summary must still be displayed so the user can interrupt if values look wrong.

Anti-Patterns

! When deft-directive-setup generates or updates USER.md or PROJECT-DEFINITION.vbrief.json, the deft_version field MUST be set to the current framework version
⊗ Generate a USER.md or PROJECT-DEFINITION.vbrief.json without including the deft_version field
⊗ Explore codebase before Phase 1 questions
⊗ Read framework files before first question
⊗ Batch multiple questions into one message — ask one at a time, interview style
⊗ Ask jargon-heavy questions to non-technical users
⊗ Ask about things inferable from codebase (Phase 2+)
⊗ Skip phases without asking
⊗ Generate files without confirming content
⊗ Present choices through a host UI that replaces the canonical numbers with alphabetic affordances or unlabeled buttons
⊗ Resolve paths relative to the skill file, AGENTS.md, or framework directory instead of the user's pwd at skill entry
⊗ Generate an authoritative PRD.md — PRD.md is a read-only export via task prd:render, never a source of truth

About

Deft Directive Setup

When to Use

Pre-Cutover Detection Guard

Detection Criteria

Action on Detection

Environment Preflight (before asking to run migration)

Prompt and Run

Greenfield Projects (No Migration Needed)

Migration safety flags

Deterministic Questions Contract

Platform Detection

Agent Behavior

Available Languages

Available Strategies

Phase 1 — User Preferences (USER.md)

USER.md Freshness Detection

Interview Rules

Question Sequence

Output Path

Template

Then

Phase 2 — Project Configuration (PROJECT-DEFINITION.vbrief.json)

Inference

Track Detection

Defaults in Agentic Mode

Interview Rules (same as Phase 1)

Question Sequence

Output Path

GitHub PR Template Scaffolding (#531)

Headless Coverage Warning — display-bound GUI entry points (#1027)

Template

Then

Follow-up: triage onboarding (#1143)

Phase 3 — Specification

Onboarding Question

⚠️ MANDATORY: Strategy Gate — Do This First

Sizing Gate (interview and yolo strategies only)

Interview Process (interview strategy)

Output — Light Path

Output — Full Path

Lifecycle Bridge to Downstream Skills (#1025)

End-of-Phase-3 Export Prompt and Render Gate

Handoff to deft-directive-build

Warp Auto-Approve Warning

Post-Interview Confirmation Gate

Anti-Patterns

Use it

Manifest

Reviews