godpowers 0.15.9 → 0.15.10

package/CHANGELOG.md

@@ -5,6 +5,22 @@ All notable changes to Godpowers will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.15.10] - 2026-05-11
+
+God Mode continuity release. Makes red verification output repair work inside
+the same autonomous arc instead of a terminal summary.
+
+### Changed
+- `/god-mode` now treats failed tests, lint, typecheck, build, or check commands
+  as repairable work and enters an autonomous repair loop before declaring the
+  arc complete.
+- Build completion now requires test, lint, and typecheck/check commands to be
+  green when those commands exist.
+- `/god-mode --yolo` now auto-runs repair loops, with Critical security findings
+  remaining the only unconditional pause.
+- Roadmap and build planning language now uses Godpowers delivery increments
+  instead of preserving imported methodology terminology.
+
 ## [0.15.9] - 2026-05-11
 
 Early design planning release. Lets UI and product-experience projects shape

package/README.md

@@ -2,7 +2,7 @@
 
 [![CI](https://github.com/aihxp/godpowers/actions/workflows/ci.yml/badge.svg)](https://github.com/aihxp/godpowers/actions/workflows/ci.yml)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
-[![Version](https://img.shields.io/badge/version-0.15.9-blue)](CHANGELOG.md)
+[![Version](https://img.shields.io/badge/version-0.15.10-blue)](CHANGELOG.md)
 [![npm](https://img.shields.io/npm/v/godpowers.svg)](https://www.npmjs.com/package/godpowers)
 
 **Ship fast. Ship right. Ship everything. Ship accountably.**
@@ -141,13 +141,18 @@ For real-world scenarios beyond greenfield:
 
 ```
 /god-mode                # Standard: pauses for real questions only
-/god-mode --yolo         # Zero pauses. Picks every default. Full send.
+/god-mode --yolo         # Zero pauses except Critical security. Repairs red checks before it stops.
 /god-mode --conservative # More checkpoints
 /god-mode --from=arch    # Resume from a specific tier
 /god-mode --audit        # Score existing artifacts. Build nothing.
 /god-mode --dry-run      # Plan everything. Build nothing.
 ```
 
+`/god-mode` is not complete when it merely writes planning artifacts. It keeps
+going through build, verification, repair, launch, and final sync. Red tests,
+typecheck, lint, build, or check output enter the repair loop instead of being
+reported as the final result.
+
 ## Architecture
 
 ### Slash Command + Specialist Agent Pattern

package/agents/god-executor.md

@@ -20,6 +20,8 @@ You receive:
 - Relevant ARCH context (only what's needed for this slice)
 - Stack DECISION (tooling)
 - The slice's dependencies (what must already exist)
+- Optional repair payload: failing command, error counts, focused diagnostics,
+  and files implicated by a previous verification run
 
 ## TDD Sequence (mandatory)
 
@@ -54,10 +56,12 @@ For every behavior in this slice:
 
 1. Run the full test suite. All tests must pass.
 2. Run the linter. All warnings resolved.
-3. Stage your changes.
-4. Return control to orchestrator with:
+3. Run typecheck/check command when present. All errors resolved.
+4. Stage your changes.
+5. Return control to orchestrator with:
    - Summary of what was implemented
    - Test results
+   - Typecheck/check results
    - Files changed
    - Ready for two-stage review
 
@@ -73,4 +77,13 @@ happen.
 - Multiple slices touched in one execution
 - Linter warnings unresolved
 - Test suite failing (any test, not just yours)
+- Typecheck/check command failing
 - Stub/placeholder code in the implementation
+
+## Repair Mode
+
+If invoked with a repair payload, stay narrowly focused on the failing command.
+Do not reopen PRD, ARCH, roadmap, or stack unless the diagnostic proves the
+artifact is stale. Fix code, config, imports, tests, generated types, or
+tooling until the command passes. If the same root failure survives 3 focused
+attempts, return the smallest human-only question needed to continue.

package/agents/god-orchestrator.md

@@ -277,17 +277,52 @@ last known hash in state.json:
 
 ## Extended Critical-Finding Gate
 
-The existing critical-finding gate now also fires on:
-- Breaking drift findings (e.g., WCAG contrast fail, ARCH container
-  responsibility violation, STACK dep version mismatch)
-- Critical impeccable findings on the diff
-- Lint errors from `lib/artifact-linter` on any artifact
-- Validator errors from `lib/have-nots-validator`
+The critical-finding gate fires on:
+- Critical security findings from god-harden-auditor
 - god-design-reviewer BLOCK verdicts
-
-All pause both default mode AND --yolo. Lint errors cannot be
-auto-resolved; they're mechanical signal that something is structurally
-wrong.
+- Breaking drift findings that would make already-written artifacts unsafe
+  to trust without human context
+- Artifact linter or have-nots errors that still fail after repair attempts
+
+Only Critical security findings always pause, including under --yolo.
+Everything else must first enter the autonomous repair loop below. A failed
+typecheck, lint, check, unit test, generated artifact lint, or have-nots pass is
+not a reason to declare the arc complete. It is work.
+
+## Autonomous Repair Loop
+
+Godpowers full-arc means: plan, build, verify, repair, ship, sync. Do not stop
+at "artifacts generated" when the repo is still red.
+
+When any mechanical verification fails:
+- tests, typecheck, lint, formatter, build, `bun run check`, `npm run check`,
+  `npm test`, `cargo test`, `go test`, or equivalent
+- artifact lint or have-nots validation
+- generated scaffold audit with fixable failures
+- launch smoke check with deterministic reproduction
+
+Do this:
+1. Record the exact failing command, counts, and highest-signal diagnostics in
+   `.godpowers/build/STATE.md` or the active tier state file.
+2. Classify the failure:
+   - `repairable`: code, config, type, lint, test, generated artifact, missing
+     dependency, bad scaffold, or stale state problem.
+   - `human-only`: product scope contradiction, credential missing, paid vendor
+     decision, legal/compliance choice, or Critical security acceptance.
+3. For `repairable`, spawn the owning agent again in repair mode with only the
+   failing diagnostics, touched files, relevant artifact excerpts, and the
+   command to re-run.
+4. Re-run the failing command after each repair attempt.
+5. Repeat until green or until the same root failure survives 3 repair attempts.
+6. If repair succeeds, continue the same `/god-mode` run. Do not hand off a
+   "next recommended delivery increment" while required verification is red.
+7. If the same root failure survives 3 attempts, pause with a precise blocker,
+   attempted fixes, and the smallest human question needed to continue.
+
+Under `--yolo`, the repair loop auto-runs. It may commit atomic repair commits
+after tests pass. If a git remote exists and the user passed an explicit push
+flag or the project intent says pushing is allowed, push after the green commit
+and then continue the arc. Pushing is not a terminal state.
 
 ## YOLO Behavior with Design + Linkage
 
@@ -297,7 +332,8 @@ wrong.
 | Impeccable install prompt | Pause | Auto-yes; log |
 | PRODUCT.md interview | Pause | Pause anyway (load-bearing brand) |
 | Design token defaults | Pause | Auto-pick impeccable defaults; log |
-| Lint errors | Pause | Pause anyway |
+| Repairable mechanical failures | Repair loop | Repair loop |
+| Lint errors after 3 repair attempts | Pause | Pause with diagnostics |
 | Lint warnings | Continue, log | Continue, log |
 | Drift (informational) | Continue | Continue |
 | Drift (breaking) | Pause | Pause anyway |
@@ -317,8 +353,9 @@ wrong.
 5. Verify their output exists on disk
 6. Run have-nots check on the artifact
 7. If pass: update PROGRESS.md, move to next sub-step
-8. If fail: spawn the agent again with the failures, OR pause for human
-9. Repeat until all tiers complete
+8. If fail and repairable: enter the autonomous repair loop
+9. If fail and human-only: pause with the smallest needed question
+10. Repeat until all tiers complete and verification is green
 ```
 
 ## Specialist Agent Routing
@@ -387,7 +424,10 @@ Move to next wave only when ALL slices in current wave are committed.
 After all waves complete:
 1. Run full test suite. All must pass.
 2. Run linter. All clean.
-3. Update PROGRESS.md: Build = done
+3. Run typecheck/check command when the package exposes one. All clean.
+4. If any verification fails, run the autonomous repair loop. Do not mark
+   Build done and do not recommend later work while verification is red.
+5. Update PROGRESS.md: Build = done
 
 CRITICAL RULES (build phase):
 - Never skip god-spec-reviewer
@@ -395,6 +435,8 @@ CRITICAL RULES (build phase):
 - Never commit without BOTH stages passing
 - Each slice gets its own atomic commit
 - Each agent gets a fresh context (defeats context rot)
+- Build cannot be `done` when test, lint, typecheck, or check commands fail
+- A release blocker is a repair target, not the final answer
 
 ## Post-Launch Transition (after Tier 3 completes)
 
@@ -420,8 +462,10 @@ This step runs regardless of flags:
 - /god-mode --conservative -> sync runs (with pause for confirmation)
 - /god-mode --with-hygiene -> sync runs PLUS hygiene check
 
-This ensures every full-arc run leaves the project in a sync'd state. The
-artifact coverage is consistent across all 14 categories.
+After sync, re-run the final verification commands. If any are red, return to
+the autonomous repair loop. This ensures every full-arc run leaves the project
+green and sync'd, not merely documented. The artifact coverage is consistent
+across all 14 categories.
 
 ### Steady-State Hand-off
 
@@ -505,7 +549,8 @@ that should never be auto-accepted. If god-harden-auditor returns Critical
 findings, --yolo does NOT skip. Pause for human resolution.
 
 This is the only --yolo carve-out. All other pauses are auto-resolved with
-the agent's documented default.
+the agent's documented default, and all repairable mechanical failures are
+handled by the autonomous repair loop before the arc can be called complete.
 
 ### Pause Format
 

package/agents/god-planner.md

@@ -1,8 +1,8 @@
 ---
 name: god-planner
 description: |
-  Build planner. Reads roadmap, selects current milestone, breaks into vertical
-  slices, identifies dependencies, groups slices into parallel waves. Each
+  Build planner. Reads roadmap, selects the current delivery increment, breaks
+  it into vertical slices, identifies dependencies, groups slices into waves. Each
   slice plan includes exact file paths, tests-first sequence, and verification
   criteria.
 
@@ -20,10 +20,11 @@ Plan the build.
 
 ## Process
 
-1. Read roadmap, identify current milestone (first non-done)
+1. Read roadmap, identify the current delivery increment (first non-done
+   Now item)
 2. Read ARCH for technical context
 3. Read stack DECISION for tooling
-4. Break milestone into **vertical slices**:
+4. Break the delivery increment into **vertical slices**:
    - Each slice delivers ONE user-visible behavior end-to-end
    - NOT "set up the database" - that's horizontal
    - YES "user can create an account" - includes DB + API + UI for that behavior
@@ -44,7 +45,7 @@ Plan the build.
 Write `.godpowers/build/PLAN.md`:
 
 ```markdown
-# Build Plan: Milestone [N]
+# Build Plan: Delivery Increment [N]
 
 ## Wave 1 (parallel)
 ### Slice 1.1: User can create an account

package/agents/god-roadmapper.md

@@ -1,8 +1,8 @@
 ---
 name: god-roadmapper
 description: |
-  Sequences work into milestones with topological dependency ordering,
-  Now/Next/Later horizons, and observable completion gates. Gated on
+  Sequences work into Godpowers delivery increments with topological dependency
+  ordering, Now/Next/Later horizons, and observable completion gates. Gated on
   Architecture.
 
   Spawned by: /god-roadmap, god-orchestrator
@@ -29,21 +29,24 @@ If `.godpowers/prep/INITIAL-FINDINGS.md` exists, read it first for repo risks,
 existing tests, docs, CI, deploy, and suggested sequencing implications.
 
 If `.godpowers/prep/IMPORTED-CONTEXT.md` exists, read its delivery signals
-before sequencing milestones. Use imported GSD, Superpowers, BMAD, or similar
-stories and plans as hypothesis-level input only.
+before sequencing work. Use imported GSD, Superpowers, BMAD, or similar
+stories and plans as hypothesis-level input only. Convert imported terminology
+into Godpowers vocabulary. Do not preserve imported methodology terminology in
+Godpowers artifacts unless the user explicitly asked for it.
 
-If DESIGN.md or PRODUCT.md exists, read them before sequencing milestones.
+If DESIGN.md or PRODUCT.md exists, read them before sequencing delivery
+increments.
 Use screens, flows, components, accessibility obligations, and product voice to
 shape vertical slices and acceptance gates.
 
 Rules:
 - Do not add roadmap features that are not grounded in the PRD.
-- Use imported milestones, stories, and done-work signals to inform ordering,
+- Use imported delivery, story, and done-work signals to inform ordering,
   dependency edges, and open questions.
 - If imported context conflicts with PRD or ARCH, PRD and ARCH win.
 - Mark any imported sequencing assumption as `[HYPOTHESIS]` until confirmed by
   Godpowers artifacts or the user.
-- Keep design-derived milestones grounded in the PRD and ARCH.
+- Keep design-derived delivery increments grounded in the PRD and ARCH.
 
 ## Process
 
@@ -51,10 +54,10 @@ Rules:
 2. List all features from PRD with their priority (MUST/SHOULD/COULD)
 3. Build dependency graph from ARCH (component A depends on component B)
 4. Topologically sort
-5. Group features into milestones:
-   - Each milestone has a clear, substitution-tested goal
-   - Each milestone has an observable completion gate
-   - Each milestone is sized: S/M/L (no day-level precision without capacity input)
+5. Group features into delivery increments:
+   - Each increment has a clear, substitution-tested goal
+   - Each increment has an observable completion gate
+   - Each increment is sized: S/M/L (no day-level precision without capacity input)
 6. Assign Now/Next/Later horizons:
    - **Now**: building right now, committed
    - **Next**: planned next, flexible
@@ -69,7 +72,7 @@ as the structural starting point. Write `.godpowers/roadmap/ROADMAP.md`:
 # Roadmap
 
 ## Now
-### Milestone 1: [substitution-tested name]
+### Delivery Increment 1: [substitution-tested name]
 - Goal: [what users can do when this ships]
 - Gate: [observable completion criteria]
 - Size: S/M/L
@@ -77,20 +80,20 @@ as the structural starting point. Write `.godpowers/roadmap/ROADMAP.md`:
 - Features: [from PRD]
 
 ## Next
-[milestones]
+[delivery increments]
 
 ## Later
-[milestones]
+[themes]
 ```
 
 ## Have-Nots
 
 Roadmap FAILS if:
-- Milestone goal passes substitution test
+- Delivery increment goal passes substitution test
 - Completion gate is not observable
 - Feature appears that is not in the PRD
-- All milestones the same size (no prioritization)
-- No dependency edges between milestones
+- All increments the same size (no prioritization)
+- No dependency edges between increments
 - Day-level precision without capacity input
 - Later section is empty (no long-term vision)
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "godpowers",
-  "version": "0.15.9",
+  "version": "0.15.10",
   "description": "AI-powered development system: 104 slash commands and 38 specialist agents that take a project from raw idea to hardened production. Runs inside Claude Code, Codex, Cursor, Windsurf, Gemini, and 10+ other AI coding tools.",
   "bin": {
     "godpowers": "./bin/install.js"

package/skills/god-build.md

@@ -51,13 +51,19 @@ Move to next wave only when current wave is fully committed.
 After all waves:
 1. Run full test suite. All pass.
 2. Run linter. All clean.
-3. Update PROGRESS.md: Build status = done
+3. Run the package's typecheck/check command when present. All pass.
+4. If any verification command fails, do not mark Build complete. Re-enter
+   repair mode with the owning agent, pass the exact failing diagnostics, rerun
+   the command, and repeat until green or until the same root failure survives
+   3 repair attempts.
+5. Update PROGRESS.md: Build status = done
 
 ## Pause Conditions
 
 Pause for user ONLY if:
 - A requirement is genuinely ambiguous (two valid implementations)
 - A test reveals a gap in PRD or ARCH that needs human resolution
+- The same mechanical failure remains after 3 focused repair attempts
 
 ## On Completion
 
@@ -70,8 +76,9 @@ Alternative: /god-deploy (set up deploy pipeline, parallel-safe)
 Both can run; /god-harden is the critical path to Launch.
 ```
 
-If more milestones remain in the roadmap, suggest re-running /god-build for
-the next milestone before moving to Tier 3.
+If more delivery increments remain in the roadmap, continue building the next
+increment before moving to Tier 3 unless the user explicitly asked to stop
+after the current increment.
 
 
 ## Locking

package/skills/god-mode.md

@@ -53,6 +53,10 @@ You are receiving a /god-mode invocation. Your job is to spawn the
    - Instruction to run `/god-design` after `/god-prd` and before `/god-arch`
      when initial findings, imported planning context, the PRD, or the
      codebase show UI or product-experience signals
+   - Instruction that a red test, typecheck, lint, build, or check command is
+     not a completed arc. It must enter the autonomous repair loop and continue
+     the same `/god-mode` run until green, except for Critical security or a
+     genuine human-only decision.
 
 6. Orchestrator runs the appropriate workflow:
    - Greenfield -> full-arc

package/skills/god-roadmap.md

@@ -1,10 +1,10 @@
 ---
 name: god-roadmap
 description: |
-  Sequence the work into milestones. Spawns the god-roadmapper agent in a
-  fresh context. Gated on Architecture.
+  Sequence the work into Godpowers delivery increments. Spawns the
+  god-roadmapper agent in a fresh context. Gated on Architecture.
 
-  Triggers on: "god roadmap", "/god-roadmap", "sequence the work", "milestone plan"
+  Triggers on: "god roadmap", "/god-roadmap", "sequence the work", "delivery plan"
 ---
 
 # /god-roadmap

package/templates/ROADMAP.md

@@ -1,11 +1,11 @@
 # Roadmap
 
-> Every milestone goal is substitution-tested. Every completion gate is
+> Every delivery increment goal is substitution-tested. Every completion gate is
 > observable. No day-level precision without capacity input.
 
 ## Now (currently building, committed)
 
-### Milestone 1: [Substitution-tested goal]
+### Delivery Increment 1: [Substitution-tested goal]
 - **Goal**: [What users can DO when this ships. Specific.]
 - **Completion gate**: [Observable criterion. Not "feels done".]
 - **Size**: S / M / L
@@ -14,20 +14,20 @@
   - [PRD requirement reference]
   - [PRD requirement reference]
 
-### Milestone 2: [Substitution-tested goal]
+### Delivery Increment 2: [Substitution-tested goal]
 - **Goal**: ...
 - **Completion gate**: ...
 - **Size**: S / M / L
-- **Depends on**: Milestone 1
+- **Depends on**: Delivery Increment 1
 - **Features**: ...
 
 ## Next (planned, flexible)
 
-### Milestone 3: [Goal]
+### Delivery Increment 3: [Goal]
 - **Goal**: ...
 - **Completion gate**: ...
 - **Size**: ...
-- **Depends on**: Milestone 2
+- **Depends on**: Delivery Increment 2
 
 ## Later (intent, not commitment)
 
@@ -38,10 +38,10 @@
 
 ## Have-Nots Checklist
 
-- [ ] Every milestone goal fails substitution test
+- [ ] Every delivery increment goal fails substitution test
 - [ ] Every completion gate is observable (not subjective)
 - [ ] No feature appears that isn't in the PRD
-- [ ] Milestones are NOT all the same size (real prioritization)
+- [ ] Delivery increments are NOT all the same size (real prioritization)
 - [ ] Dependency edges are explicit
 - [ ] No day-level precision without capacity input
 - [ ] Later section has at least one theme

package/workflows/full-arc.yaml

@@ -58,6 +58,9 @@ jobs:
         - uses: god-quality-reviewer@^1.0.0
           on-fail: rollback
     on-pass: commit-atomic
+    on-fail: repair-loop
+    verification:
+      required: [test, lint, typecheck-or-check]
 
   deploy:
     tier: 3