devflow-kit 1.5.0 → 1.6.1

package/shared/skills/implementation-orchestration/SKILL.md

@@ -0,0 +1,92 @@
+---
+name: implementation-orchestration
+description: Agent orchestration for IMPLEMENT intent — pre-flight, Coder, quality gates
+user-invocable: false
+allowed-tools: Read, Grep, Glob, Bash, Task, AskUserQuestion
+---
+
+# Implementation Orchestration
+
+Agent pipeline for IMPLEMENT intent in ambient ORCHESTRATED mode. Pre-flight checks, plan synthesis, Coder execution, and quality gates.
+
+This is a lightweight variant of `/implement` for ambient ORCHESTRATED mode. Excluded: strategy selection (single/sequential/parallel Coders), retry loops, PR creation, knowledge loading.
+
+## Iron Law
+
+> **QUALITY GATES ARE NON-NEGOTIABLE**
+>
+> Every Coder output passes through Validator → Simplifier → Scrutinizer → re-Validate → Shepherd.
+> Skipping a gate because "it looks fine" is never acceptable. The pipeline runs to completion
+> or halts on failure — there is no shortcut.
+
+---
+
+## Phase 1: Pre-flight — Branch Safety
+
+Detect branch type before spawning Coder:
+
+- **Work branches** (`feat/`, `fix/`, `chore/`, `refactor/`, `docs/` prefix): proceed on current branch.
+- **Protected branches** (`main`, `master`, `develop`, `release/*`, `staging`, `production`): ask user via AskUserQuestion with 2-3 suggested branch names following `{type}/{ticket}-{slug}` convention. Include ticket number if available from conversation context.
+- **If user declines branch creation**: proceed on the protected branch. Respect the user's choice.
+
+## Phase 2: Plan Synthesis
+
+Synthesize conversation context into a structured EXECUTION_PLAN for Coder:
+
+- **If a plan exists** in conversation context (from plan mode — accepted in-session or injected after "accept and clear") → use the plan as-is.
+- **Otherwise** → synthesize from conversation: what to build, files/modules affected, constraints, decisions made during discussion.
+
+Format as structured markdown with: Goal, Steps, Files, Constraints, Decisions.
+
+## Phase 3: Coder Execution
+
+Record git SHA before first Coder: `git rev-parse HEAD`
+
+Spawn `Task(subagent_type="Coder")` with input variables:
+- **TASK_ID**: Generated from timestamp (e.g., `task-2026-03-19_1430`)
+- **TASK_DESCRIPTION**: From conversation context
+- **BASE_BRANCH**: Current branch (or newly created branch from Phase 1)
+- **EXECUTION_PLAN**: From Phase 2
+- **PATTERNS**: Codebase patterns from conversation context
+- **CREATE_PR**: `false` (commit only, no push)
+- **DOMAIN**: Inferred from files in scope (`backend`, `frontend`, `tests`, `fullstack`)
+
+**Execution strategy**: Single sequential Coder by default. Parallel Coders only when tasks are self-contained — zero shared contracts, no integration points, different files/modules with no imports between them.
+
+If Coder returns **BLOCKED**, halt the pipeline and report to user.
+
+## Phase 4: FILES_CHANGED Detection
+
+After Coder completes, detect changed files:
+
+```bash
+git diff --name-only {starting_sha}...HEAD
+```
+
+Pass FILES_CHANGED to all quality gate agents.
+
+## Phase 5: Quality Gates
+
+Run sequentially — each gate must pass before the next:
+
+1. `Task(subagent_type="Validator")` (build + typecheck + lint + tests) — retry up to 2× on failure (Coder fixes between retries)
+2. `Task(subagent_type="Simplifier")` — code clarity and maintainability pass on FILES_CHANGED
+3. `Task(subagent_type="Scrutinizer")` — 9-pillar quality evaluation on FILES_CHANGED
+4. `Task(subagent_type="Validator")` (re-validate after Simplifier/Scrutinizer changes)
+5. `Task(subagent_type="Shepherd")` — verify implementation matches original request — retry up to 2× if misalignment found
+
+If any gate exhausts retries, halt pipeline and report what passed and what failed.
+
+## Phase 6: Completion
+
+Report results:
+- Commits created (from Coder)
+- Files changed
+- Quality gate results (pass/fail per gate)
+- No push — user decides when to push
+
+## Error Handling
+
+- **Coder BLOCKED**: Halt immediately, report blocker to user
+- **Validator fails after retries**: Report specific failures, halt pipeline
+- **Shepherd misalignment after retries**: Report misalignment details, let user decide next steps

package/shared/skills/knowledge-persistence/SKILL.md

@@ -0,0 +1,128 @@
+---
+name: knowledge-persistence
+description: >-
+  This skill should be used when recording architectural decisions or pitfalls
+  to project knowledge files, or when loading prior decisions and known pitfalls
+  for context during investigation, specification, or review.
+user-invocable: false
+allowed-tools: Read, Write, Bash
+---
+
+# Knowledge Persistence
+
+Record architectural decisions and pitfalls to `.memory/knowledge/` files. This is the single source of truth for the extraction procedure — commands reference this skill instead of inlining the steps.
+
+## Iron Law
+
+> **SINGLE SOURCE OF TRUTH**
+>
+> All knowledge extraction follows this procedure exactly. Commands never inline
+> their own extraction steps — they read this skill and follow it.
+
+---
+
+## File Locations
+
+```
+.memory/knowledge/
+├── decisions.md    # ADR entries (append-only)
+└── pitfalls.md     # PF entries (area-specific gotchas)
+```
+
+## File Formats
+
+### decisions.md (ADR entries)
+
+**Template header** (create if file missing):
+```
+<!-- TL;DR: 0 decisions. Key: -->
+# Architectural Decisions
+
+Append-only. Status changes allowed; deletions prohibited.
+```
+
+**Entry format**:
+```markdown
+## ADR-{NNN}: {Title}
+
+- **Date**: {YYYY-MM-DD}
+- **Status**: Accepted
+- **Context**: {Why this decision was needed}
+- **Decision**: {What was decided}
+- **Consequences**: {Tradeoffs and implications}
+- **Source**: {command and identifier, e.g. `/implement TASK-123`}
+```
+
+### pitfalls.md (PF entries)
+
+**Template header** (create if file missing):
+```
+<!-- TL;DR: 0 pitfalls. Key: -->
+# Known Pitfalls
+
+Area-specific gotchas, fragile areas, and past bugs.
+```
+
+**Entry format**:
+```markdown
+## PF-{NNN}: {Short description}
+
+- **Area**: {file paths or module names}
+- **Issue**: {What goes wrong}
+- **Impact**: {Consequences if hit}
+- **Resolution**: {How to fix or avoid}
+- **Source**: {command and identifier, e.g. `/code-review branch-name`}
+```
+
+---
+
+## Extraction Procedure
+
+Follow these steps when recording decisions or pitfalls:
+
+1. **Read** the target file (`.memory/knowledge/decisions.md` or `.memory/knowledge/pitfalls.md`). If it doesn't exist, create it with the template header above.
+2. **Check capacity** — count `## ADR-` or `## PF-` headings. If >=50, log "Knowledge base at capacity — skipping new entry" and stop.
+3. **Find next ID** — find highest NNN via regex (`/^## ADR-(\d+)/` or `/^## PF-(\d+)/`), default to 0. Increment by 1.
+4. **Deduplicate** (pitfalls only) — skip if an entry with the same Area + Issue already exists.
+5. **Append** the new entry using the format above.
+6. **Update TL;DR** — rewrite the `<!-- TL;DR: ... -->` comment on line 1 to reflect the new count and key topics.
+
+## Lock Protocol
+
+When writing, use a mkdir-based lock:
+- Lock path: `.memory/.knowledge.lock`
+- Timeout: 30 seconds (fail if lock not acquired)
+- Stale recovery: if lock directory is >60 seconds old, remove it and retry
+- Release lock after write completes (remove lock directory)
+
+## Loading Knowledge for Context
+
+When a command needs prior knowledge as input (not recording):
+
+1. Read `.memory/knowledge/decisions.md` if it exists
+2. Read `.memory/knowledge/pitfalls.md` if it exists
+3. Pass content as context to downstream agents — prior decisions constrain scope, known pitfalls inform investigation
+
+If neither file exists, skip silently. No error, no empty-file creation.
+
+## Operation Budget
+
+Recording: do inline (no agent spawn), 2-3 Read/Write operations total.
+Loading: 1-2 Read operations, pass as context string.
+
+---
+
+## Extended References
+
+For entry examples and status lifecycle details:
+- `references/examples.md` - Full decision and pitfall entry examples
+
+---
+
+## Success Criteria
+
+- [ ] Entry appended with correct sequential ID
+- [ ] No duplicate pitfalls (same Area + Issue)
+- [ ] TL;DR comment updated with current count
+- [ ] Lock acquired before write, released after
+- [ ] Capacity limit (50) respected

package/shared/skills/knowledge-persistence/references/examples.md

@@ -0,0 +1,44 @@
+# Knowledge Persistence Examples
+
+## Decision Entry Example
+
+```markdown
+## ADR-001: Use mkdir-based locks for concurrent session serialization
+
+- **Date**: 2026-03-03
+- **Status**: Accepted
+- **Context**: Multiple Claude Code sessions can run on the same project simultaneously (different terminals, SSH, etc.). Memory writes must serialize to prevent corruption.
+- **Decision**: Use `mkdir` as an atomic lock primitive. Lock directory at `.memory/.knowledge.lock`. 30-second timeout with 60-second stale recovery.
+- **Consequences**: Simple, cross-platform, no external dependencies. Cannot detect holder PID if lock is stale — relies on age-based recovery. Sufficient for low-contention writes.
+- **Source**: `/implement #99`
+```
+
+## Pitfall Entry Example
+
+```markdown
+## PF-001: Orphaned teams variants silently skipped
+
+- **Area**: plugins/devflow-*/commands/*-teams.md, src/cli/installer
+- **Issue**: The installer iterates base `.md` files and looks up matching `-teams.md` variants. A `-teams.md` file without a corresponding base `.md` is silently ignored during installation.
+- **Impact**: Teams variant appears committed but never installs. Users on `--teams` mode silently get no command.
+- **Resolution**: Always create the base `.md` file first. CI should validate that every `-teams.md` has a matching base file.
+- **Source**: `/code-review feat/agent-teams`
+```
+
+## Status Lifecycle (Decisions Only)
+
+Decisions support status transitions:
+- `Accepted` — current, in effect
+- `Superseded by ADR-NNN` — replaced by a newer decision
+- `Deprecated` — no longer relevant, kept for history
+
+Pitfalls have no status field — they remain until manually removed.
+
+## Deduplication Logic (Pitfalls Only)
+
+Before appending a new pitfall, check existing entries:
+1. Extract `Area` and `Issue` from the new entry
+2. Compare against all existing `PF-*` entries
+3. If both Area AND Issue match an existing entry (case-insensitive substring), skip
+
+This prevents recording the same gotcha from multiple review cycles.

package/shared/skills/plan-orchestration/SKILL.md

@@ -0,0 +1,71 @@
+---
+name: plan-orchestration
+description: Agent orchestration for PLAN intent — codebase orientation, design exploration, gap validation
+user-invocable: false
+allowed-tools: Read, Grep, Glob, Bash, Task, AskUserQuestion
+---
+
+# Plan Orchestration
+
+Agent pipeline for PLAN intent in ambient ORCHESTRATED mode. Codebase orientation, targeted exploration, architecture design, and gap validation.
+
+This is a lightweight variant of the Plan phase in `/implement` for ambient ORCHESTRATED mode.
+
+## Iron Law
+
+> **PLANS WITHOUT CODEBASE GROUNDING ARE FANTASIES**
+>
+> Orient before architecting. Every design decision must reference existing patterns,
+> real file structures, and actual integration points. A plan that ignores the codebase
+> will fail on contact with implementation.
+
+---
+
+## Phase 1: Orient
+
+Spawn `Task(subagent_type="Skimmer")` to get codebase overview relevant to the planning question:
+
+- Existing patterns and conventions in the affected area
+- File structure and module boundaries
+- Test patterns and coverage approach
+- Related prior implementations (similar features, analogous patterns)
+
+## Phase 2: Explore
+
+Based on Skimmer findings, spawn 2-3 `Task(subagent_type="Explore")` agents **in a single message** (parallel execution):
+
+- **Integration explorer**: Examine integration points — APIs, shared types, module boundaries the plan must respect
+- **Pattern explorer**: Find existing implementations of similar features to follow as templates
+- **Constraint explorer**: Identify constraints — test infrastructure, build system, CI requirements, deployment concerns
+
+Adjust explorer focus based on the specific planning question.
+
+## Phase 3: Design
+
+Spawn `Task(subagent_type="Plan")` with combined Skimmer + Explore findings:
+
+- Design implementation approach with file-level specificity
+- Reference existing patterns discovered in Phase 1-2
+- Include: architecture decisions, file changes, new files needed, test strategy
+- Flag any areas where existing patterns conflict with the proposed approach
+
+## Phase 4: Validate
+
+Main session reviews the plan for:
+
+- **Gaps**: Missing files, unhandled edge cases, integration points not addressed
+- **Risks**: Areas where the plan deviates from existing patterns, potential regressions
+- **Ambiguities**: Design choices that need user input
+
+Present plan to user with identified risks. Use AskUserQuestion for any ambiguous design choices.
+
+## Output
+
+Structured plan ready to feed into IMPLEMENT/ORCHESTRATED if user proceeds:
+
+- Goal and scope
+- Architecture decisions with rationale
+- File-level change list (create/modify/delete)
+- Test strategy
+- Risks and mitigations
+- Open questions (if any)

package/shared/skills/test-driven-development/SKILL.md

@@ -91,7 +91,7 @@ See `references/rationalization-prevention.md` for extended examples with code.
 
 ## Process Enforcement
 
-When implementing any feature under ambient BUILD/GUIDED:
+When implementing any feature under ambient IMPLEMENT/GUIDED or IMPLEMENT/ORCHESTRATED:
 
 1. **Identify the first behavior** — What is the simplest thing this feature must do?
 2. **Write the test** — Describe that behavior as a failing test
@@ -130,7 +130,8 @@ When skipping TDD, never rationalize. State clearly: "Skipping TDD because: [spe
 
 ## Integration with Ambient Mode
 
-- **BUILD/GUIDED** → TDD enforced. Every new function/method gets test-first treatment.
-- **BUILD/QUICK** → TDD skipped (trivial single-file edit).
-- **BUILD/ELEVATE** → TDD mentioned in nudge toward `/implement`.
-- **DEBUG/GUIDED** → TDD applies to the fix: write a test that reproduces the bug first, then fix.
+- **IMPLEMENT/GUIDED** → TDD enforced in main session. Write the failing test before production code. Skill loaded directly.
+- **IMPLEMENT/ORCHESTRATED** → TDD enforced via Coder agent (skill in Coder frontmatter). Every implementation gets test-first treatment.
+- **IMPLEMENT/QUICK** → TDD skipped (trivial single-file edit).
+- **DEBUG/GUIDED** → TDD applies to the fix in main session: write a test that reproduces the bug first, then fix.
+- **DEBUG/ORCHESTRATED** → TDD applies to the fix: write a test that reproduces the bug first, then fix.

package/plugins/devflow-ambient/commands/ambient.md

@@ -1,110 +0,0 @@
----
-description: Ambient mode — classify intent and auto-load relevant skills for any prompt
----
-
-# Ambient Command
-
-Classify user intent and auto-load relevant skills. No agents spawned — enhances the main session only.
-
-## Usage
-
-```
-/ambient <prompt>           Classify and respond with skill enforcement
-/ambient                    Show usage
-```
-
-## Phases
-
-### Phase 1: Load Router
-
-Read the `ambient-router` skill:
-- `~/.claude/skills/ambient-router/SKILL.md`
-
-### Phase 2: Classify
-
-Apply the ambient-router classification to `$ARGUMENTS`:
-
-1. **Intent:** BUILD | DEBUG | REVIEW | PLAN | EXPLORE | CHAT
-2. **Depth:** QUICK | GUIDED | ELEVATE
-
-If no arguments provided, output:
-
-```
-## Ambient Mode
-
-Classify intent and auto-load relevant skills.
-
-Usage: /ambient <your prompt>
-
-Examples:
-  /ambient add a login form          → BUILD/GUIDED (loads TDD + implementation-patterns)
-  /ambient fix the auth error        → DEBUG/GUIDED (loads test-patterns + core-patterns)
-  /ambient where is the config?      → EXPLORE/QUICK (responds normally)
-  /ambient refactor the auth system  → BUILD/ELEVATE (suggests /implement)
-
-Always-on: devflow ambient --enable
-```
-
-Then stop.
-
-### Phase 3: State Classification
-
-- **QUICK:** Skip this phase entirely. Respond directly in Phase 4.
-- **GUIDED:** Output one line: `Ambient: {INTENT}/{DEPTH}. Loading: {skill1}, {skill2}.`
-- **ELEVATE:** Skip — recommendation happens in Phase 4.
-
-### Phase 4: Apply
-
-**QUICK:**
-Respond to the user's prompt normally. Zero skill loading. Zero overhead.
-
-**GUIDED:**
-Read the selected skills based on the ambient-router's skill selection matrix:
-
-| Intent | Primary Skills | Secondary (conditional) |
-|--------|---------------|------------------------|
-| BUILD | test-driven-development, implementation-patterns | typescript (.ts), react (.tsx), frontend-design (CSS/UI), input-validation (forms/API), security-patterns (auth/crypto) |
-| DEBUG | test-patterns, core-patterns | git-safety (if git ops) |
-| REVIEW | self-review, core-patterns | test-patterns |
-| PLAN | implementation-patterns | core-patterns |
-
-Read up to 3 skills from `~/.claude/skills/{name}/SKILL.md`. Apply their patterns and constraints when responding to the user's prompt.
-
-For BUILD intent: enforce RED-GREEN-REFACTOR from test-driven-development. Write failing tests before production code.
-
-**ELEVATE:**
-Respond to the user's prompt with your best effort, then append:
-
-> This task spans multiple files/systems. Consider `/implement` for full lifecycle management (exploration → planning → implementation → review).
-
-## Architecture
-
-```
-/ambient <prompt> (main session, no agents)
-│
-├─ Phase 1: Load ambient-router skill
-├─ Phase 2: Classify intent + depth
-├─ Phase 3: State classification (GUIDED only)
-└─ Phase 4: Apply
-   ├─ QUICK → respond directly
-   ├─ GUIDED → load 2-3 skills, apply patterns, respond
-   └─ ELEVATE → respond + workflow nudge
-```
-
-## Edge Cases
-
-| Case | Handling |
-|------|----------|
-| No arguments | Show usage and stop |
-| Single word ("help") | Classify — likely CHAT/QUICK |
-| Prompt references `/implement` etc. | Classify as normal — user chose /ambient intentionally |
-| Mixed intent ("fix and add test") | Use higher-overhead intent (BUILD > DEBUG) |
-| User says "no enforcement" | Respect immediately — treat as QUICK |
-
-## Principles
-
-1. **No agents** — Ambient enhances the main session, never spawns subagents
-2. **Proportional** — QUICK gets zero overhead, GUIDED gets 2-3 skills, ELEVATE gets a nudge
-3. **Transparent** — State classification for GUIDED/ELEVATE, silent for QUICK
-4. **Respectful** — Never over-classify; when in doubt, go one tier lower
-5. **TDD for BUILD** — GUIDED depth BUILD tasks enforce test-first workflow