ma-agents 3.4.8 → 3.5.0

package/.opencode/skills/devops-sign-docker-image/SKILL.md

@@ -1,24 +0,0 @@
----
-name: devops-sign-docker-image
-description: Cryptographically sign Docker images using certificates and Cosign
-type: skill
-triggers:
-  - "sign docker image"
-  - "docker image signing"
----
-
-# sign-docker-image.md
-# Docker Image Signing Workflow
-
-This workflow guides the DevOps agent through the process of cryptographically signing a Docker image.
-
-## Instructions
-1.  **Select Image**: Identify the image to sign.
-2.  **Get Digest**: Retrieve the immutable digest: `docker inspect --format='{{index .RepoDigests 0}}' {image_name}`.
-3.  **Prepare Certificate**: Locate the certificate file provided by the user.
-4.  **Execute Signing**:
-    - Use the `docker-image-signing` skill.
-    - Path: `skills/docker-image-signing/scripts/sign-image.sh`
-    - Run: `bash skills/docker-image-signing/scripts/sign-image.sh {image_digest} {cert_file} {key_file}`
-5.  **Verify**: Confirm the signature using `cosign verify`.
-6.  **Report**: provide the signed image reference to the user.

package/.opencode/skills/devops-sign-docker-image/bmad-skill-manifest.yaml

@@ -1,3 +0,0 @@
-type: skill
-name: devops-sign-docker-image
-module: ma-skills

package/.opencode/skills/generate-backlog/SKILL.md

@@ -1,183 +0,0 @@
----
-name: generate-backlog
-description: Generate or refresh a flat backlog.yaml from epics and bug stories, preserving existing priority and sprint assignments
-type: skill
-triggers:
-  - "generate backlog"
-  - "refresh backlog"
----
-
-# Generate Backlog Workflow
-
-Generate or refresh a flat `backlog.yaml` from epic stories and bug reports. Preserves existing priority and sprint assignments, appends new items, removes stale items, and re-numbers priorities 1..N.
-
-<workflow>
-
-<step n="1" goal="Load sources">
-  <action>Read `_bmad-output/planning-artifacts/epics.md` — this is the master list of epics with their stories.</action>
-  <check if="epics.md does not exist">
-    <output>**Error:** `_bmad-output/planning-artifacts/epics.md` not found. Cannot generate backlog without an epics file.</output>
-    <action>Exit workflow</action>
-  </check>
-
-  <action>Glob `_bmad-output/implementation-artifacts/bug-*.md` to discover all bug story files.</action>
-  <action>Store the list of bug files as {{bug_files}} (may be empty — zero bugs is valid).</action>
-
-  <action>Read `_bmad-output/implementation-artifacts/sprint-status.yaml` if it exists — this provides the current status of each story (done, in-progress, backlog, etc.).</action>
-  <action>Store the status map as {{status_map}} (keyed by story id). If the file does not exist, use an empty map.</action>
-
-  <action>Read `_bmad-output/implementation-artifacts/backlog.yaml` if it exists — this is the existing backlog to merge with.</action>
-  <action>Store the existing backlog items as {{existing_items}} (list). If the file does not exist, use an empty list.</action>
-
-  <output>
-**Sources loaded:**
-- epics.md: found
-- Bug files: {{bug_files | length}} file(s)
-- sprint-status.yaml: {{status_map exists ? "found" : "not found (empty status)"}}
-- Existing backlog.yaml: {{existing_items | length}} item(s)
-  </output>
-</step>
-
-<step n="2" goal="Extract stories from epics">
-  <action>Parse `epics.md` to extract every story from every epic. Each epic is a level-2 heading (`## Epic N: ...`) and each story within it is a level-3 heading (`### Story N.M: ...`) or a list item following the pattern `- **Story N.M:** Title` or `- N.M: Title`.</action>
-
-  <action>For each story found:
-    - Derive {{story_id}} as a kebab-case slug: `"{epic_number}-{story_number}-{title_slug}"` (e.g., `"5-5-explicit-parameter-passing"`)
-      - Title slug: lowercase, replace spaces/underscores with hyphens, remove non-alphanumeric/non-hyphen characters, collapse consecutive hyphens, trim leading/trailing hyphens
-    - Set {{epic}} to the epic number (integer)
-    - Set {{title}} to the story title text
-    - Look up status from {{status_map}} using the story identifier. If found and status is `done`, skip this story (exclude from backlog). If found, use that status. If not found, default to `backlog`.
-    - Set {{type}} = `story`
-    - Set {{severity}} = `null`
-    - Set {{sprint}} = `null`
-    - Set {{priority}} = `null` (will be assigned during merge/sort)
-  </action>
-
-  <action>Store the extracted stories as {{extracted_stories}} list.</action>
-  <output>Extracted {{extracted_stories | length}} active stories from epics (done items excluded).</output>
-</step>
-
-<step n="3" goal="Extract bugs">
-  <action>For each file in {{bug_files}}:
-    - Read the file and parse the YAML frontmatter (between `---` delimiters)
-    - Extract `title`, `severity`, and `status` from frontmatter
-    - Derive {{bug_id}} from the **filename**, not the title: strip the `bug-` prefix and `.md` extension, then uppercase to `"BUG-{slug}"` (e.g., `bug-login-crash.md` → `BUG-login-crash`)
-    - Use `status` from frontmatter directly (bugs track their own status). If `status` is missing, default to `backlog`. If status is `done`, skip this bug (exclude from backlog).
-    - Build a backlog item:
-      - id: {{bug_id}}
-      - type: `bug`
-      - epic: `null`
-      - title: {{title}}
-      - priority: `null` (assigned during merge/sort)
-      - status: {{frontmatter_status or "backlog"}}
-      - sprint: `null`
-      - severity: {{severity}}
-  </action>
-
-  <action>Store the extracted bugs as {{extracted_bugs}} list.</action>
-  <output>Extracted {{extracted_bugs | length}} active bugs (done items excluded).</output>
-</step>
-
-<step n="4" goal="Merge with existing backlog">
-  <action>Combine {{extracted_stories}} and {{extracted_bugs}} into {{new_items}} list.</action>
-
-  <action>Build a lookup map from {{existing_items}} keyed by `id`.</action>
-
-  <action>For each item in {{new_items}}:
-    - If an item with the same `id` exists in {{existing_items}}:
-      - **Preserve** the existing `priority` value
-      - **Preserve** the existing `sprint` assignment
-      - **Update** `status` from the freshly extracted value (which may reflect sprint-status.yaml changes)
-      - **Update** `title`, `type`, `epic`, `severity` from the freshly extracted values (source of truth)
-    - If no matching item exists in {{existing_items}}:
-      - This is a new item — keep all freshly extracted values
-      - Set `priority` to `null` (will be placed at the end during sort)
-  </action>
-
-  <action>Identify stale items: any item in {{existing_items}} whose `id` does NOT appear in {{new_items}}. These items no longer exist in the source files and should be removed.</action>
-
-  <action>Store the final merged list as {{merged_items}}.</action>
-  <action>Store the count of new items added as {{new_count}}.</action>
-  <action>Store the count of stale items removed as {{stale_count}}.</action>
-  <action>Store the count of preserved items as {{preserved_count}}.</action>
-
-  <output>
-**Merge results:**
-- Preserved (existing): {{preserved_count}}
-- New items added: {{new_count}}
-- Stale items removed: {{stale_count}}
-  </output>
-</step>
-
-<step n="5" goal="Write backlog.yaml">
-  <action>Sort {{merged_items}} for final output:
-    1. Items WITH an existing numeric `priority` come first, sorted ascending by priority
-    2. Items WITHOUT a priority (`null`) come after, sorted by:
-       a. `type`: bugs before stories (bugs get priority attention)
-       b. `severity` for bugs: critical > high > medium > low
-       c. `epic` number ascending for stories (lower epics first)
-       d. `id` alphabetically as final tiebreaker
-  </action>
-
-  <action>Re-number priorities sequentially 1..N across all items in the sorted list.</action>
-
-  <action>Get current ISO timestamp as {{generated_at}}.</action>
-
-  <action>Write `_bmad-output/implementation-artifacts/backlog.yaml` with this structure:
-
-```yaml
-# Flat Backlog — auto-generated by generate-backlog
-# Generated: {{generated_at}}
-# Total items: {{merged_items | length}}
-#
-# Schema per item:
-#   id, type, epic, title, priority, status, sprint, severity
-
-backlog:
-  - id: "5-5-explicit-parameter-passing"
-    type: story
-    epic: 5
-    title: "Explicit Parameter Passing"
-    priority: 1
-    status: backlog
-    sprint: null
-    severity: null
-  # ... remaining items ...
-```
-
-  **YAML output rules:**
-  - Always quote `id` and `title` fields (they contain hyphens and special characters)
-  - `epic` is an integer for stories, `null` for bugs
-  - `sprint` is a string like `"sprint-3"` if assigned, or `null`
-  - `severity` is a string (`critical`, `high`, `medium`, `low`) for bugs, `null` for stories
-  - `status` is one of: `backlog`, `ready-for-dev`, `in-progress`, `review` (done items excluded)
-  - `priority` is always a positive integer (1 = highest priority)
-  - Use 2-space indentation for YAML
-  </action>
-
-  <output>Backlog written to `_bmad-output/implementation-artifacts/backlog.yaml` with {{merged_items | length}} items.</output>
-</step>
-
-<step n="6" goal="Report summary">
-  <output>
-## Backlog Generation Complete
-
-| Metric              | Count |
-|----------------------|-------|
-| Total items          | {{merged_items | length}} |
-| Stories              | {{story_count}} |
-| Bugs                 | {{bug_count}} |
-| New items added      | {{new_count}} |
-| Stale items removed  | {{stale_count}} |
-| Items preserved      | {{preserved_count}} |
-
-**Output:** `_bmad-output/implementation-artifacts/backlog.yaml`
-
-**Next Steps:**
-- Review priorities — new items are appended at the end; reorder as needed
-- Use `/add-to-sprint` to assign high-priority items to the current sprint
-- Use `/sprint-status-view` to see sprint progress
-  </output>
-</step>
-
-</workflow>

package/.opencode/skills/generate-backlog/bmad-skill-manifest.yaml

@@ -1,3 +0,0 @@
-type: skill
-name: generate-backlog
-module: ma-skills

package/.opencode/skills/git-workflow-skill/SKILL.md

@@ -1,194 +0,0 @@
----
-name: Git Workflow
-description: MANDATORY worktree-based workflow for ALL file-changing activities. Enforces isolated feature branches, conventional commits, and PR-based merging.
----
-# Mandatory Usage Policy (CRITICAL)
-
-**This skill is MANDATORY for all activities that modify tracked files in this repository.**
-
-1.  **Before starting ANY work** that changes source code, config files, or documentation: You MUST run `./scripts/start-feature.sh` (or the equivalent `git worktree add`) to create an isolated feature branch and worktree.
-2.  **During development**: You MUST perform all file-writing activities inside the newly created worktree directory.
-3.  **After completing work**: You MUST run `./scripts/finish-feature.sh` (or the equivalent rebase/push/PR/cleanup sequence) to integrate your changes and clean up the worktree.
-
-**Failure to follow this workflow is a violation of repository standards.**
-
----
-
-# Git Workflow (Worktree-Based)
-
-Multi-agent parallel development using git worktrees. Each agent gets an isolated working directory — no branch switching, no conflicts between agents.
-
-## Bundled Tools
-
-### Scripts (scripts/)
-| Script | Purpose |
-|--------|---------|
-| `start-feature.sh <type> <desc>` | Create worktree + feature branch from dev |
-| `finish-feature.sh [path]` | Rebase, push, create PR, remove worktree |
-| `validate-workflow.sh [--list]` | Check workflow compliance / list worktrees |
-| `install-hooks.sh` | Install git hooks to repo |
-
-### Git Hooks (hooks/)
-| Hook | Purpose |
-|------|---------|
-| `pre-commit` | Blocks commits on dev/main (shared across worktrees) |
-| `commit-msg` | Validates conventional commit format |
-| `prepare-commit-msg` | Adds commit message template |
-
-**Install hooks:** Run `scripts/install-hooks.sh` in target repository.
-
-### Requirements
-- Git (2.15+ for worktree support)
-- GitHub CLI (`gh`) - Required for automatic PR creation
-  ```bash
-  # macOS
-  brew install gh
-  # Windows
-  winget install --id GitHub.cli
-  # Linux
-  apt install gh
-
-  # Authenticate
-  gh auth login
-  ```
-
----
-
-## Mandatory Pre-Change Check
-
-Before making ANY change to repository files (that are not in .gitignore):
-
-1. Check if files to be changed are tracked:
-   ```bash
-   git check-ignore -v <file-path>
-   ```
-   If file is ignored → proceed without this workflow
-   If file is tracked → continue with steps below
-
-2. Verify `dev` branch exists:
-   ```bash
-   git fetch origin
-   git branch -a | grep -E '(^|\/)dev$'
-   ```
-3. If `dev` does not exist → **STOP and ask user** to create it or specify base branch
-
-## Multi-Agent Development Flow
-
-```
-Main repo (stays on dev) ─── always clean, never modified directly
-  │
-  ├── .worktrees/auth-system/     ← Agent 1: feature/auth-system
-  │     └── (isolated copy of repo, own branch, own changes)
-  │
-  ├── .worktrees/api-refactor/    ← Agent 2: bugfix/api-refactor
-  │     └── (isolated copy of repo, own branch, own changes)
-  │
-  └── .worktrees/update-docs/     ← Agent 3: chore/update-docs
-        └── (isolated copy of repo, own branch, own changes)
-
-Each worktree:
-  1. Has its own directory and branch
-  2. Shares git history with the main repo
-  3. Can commit, push, and create PRs independently
-  4. Gets cleaned up automatically after finish-feature.sh
-```
-
-### Starting Work
-
-**Using script (recommended):**
-```bash
-./scripts/start-feature.sh feature my-feature-name
-# Output: .worktrees/my-feature-name/ created
-cd .worktrees/my-feature-name/
-```
-
-**Manual:**
-```bash
-git worktree add .worktrees/<desc> -b feature/<desc> origin/dev
-cd .worktrees/<desc>
-```
-
-### Branch Naming
-
-| Type | Use Case |
-|------|----------|
-| `feature/<desc>` | New functionality |
-| `bugfix/<desc>` | Bug fixes |
-| `hotfix/<desc>` | Urgent production fixes |
-| `chore/<desc>` | Maintenance tasks |
-
-### Completing Work
-
-**Using script (recommended):**
-```bash
-# Run from inside the worktree
-./scripts/finish-feature.sh
-```
-
-This script will:
-1. Validate you're on a feature branch
-2. Validate commit messages (conventional commits)
-3. Rebase on latest `dev` branch
-4. Push the feature branch
-5. **Automatically create a PR to `dev`** (requires `gh` CLI)
-6. **Remove the worktree** (auto-cleanup)
-
-**Note:** If `gh` CLI is not installed, the script will skip PR creation. The branch will still be pushed.
-
-**Manual:**
-```bash
-# From inside the worktree
-git fetch origin dev
-git rebase origin/dev
-git push -u origin <branch-name>
-gh pr create --base dev --head <branch-name>
-
-# Clean up from main repo
-cd /path/to/main/repo
-git worktree remove .worktrees/<desc>
-```
-
-### Managing Worktrees
-
-```bash
-# List all active worktrees
-git worktree list
-# or
-./scripts/validate-workflow.sh --list
-
-# Remove a specific worktree
-git worktree remove .worktrees/<desc>
-
-# Clean up stale references
-git worktree prune
-
-# Validate workflow compliance
-./scripts/validate-workflow.sh
-```
-
-## Commit Format (Conventional Commits)
-
-```
-<type>(<scope>): <description>
-
-[optional body]
-```
-
-**Types:** `feat`, `fix`, `chore`, `docs`, `refactor`, `test`
-
-**Examples:**
-- `feat(auth): add OAuth2 flow support`
-- `fix(api): handle null response from endpoint`
-- `chore(deps): update Helm chart dependencies`
-- `docs(readme): add deployment instructions`
-
-## Critical Rules
-
-- **NEVER** modify tracked files without being in a worktree on a feature branch.
-- **NEVER** commit directly to `dev` or `main`.
-- **ALWAYS** check if a file is gitignored before applying this workflow.
-- **ALWAYS** create a worktree **BEFORE** calling any file-writing tools (write_to_file, replace_file_content, etc.).
-- **ALWAYS** rebase on `dev` before pushing.
-- **ALWAYS** use conventional commit format.
-- **ALWAYS** add `.worktrees/` to your `.gitignore`.
-- **ALWAYS** finish the workflow by creating a PR and removing the worktree once work is verified.

package/.opencode/skills/git-workflow-skill/hooks/commit-msg

@@ -1,61 +0,0 @@
-#!/bin/bash
-# commit-msg hook - Validate conventional commit format
-# Install: cp this file to .git/hooks/commit-msg && chmod +x .git/hooks/commit-msg
-
-RED='\033[0;31m'
-GREEN='\033[0;32m'
-YELLOW='\033[1;33m'
-NC='\033[0m'
-
-COMMIT_MSG_FILE="$1"
-COMMIT_MSG=$(cat "$COMMIT_MSG_FILE")
-
-# Skip merge commits
-if echo "$COMMIT_MSG" | grep -qE '^Merge '; then
-    exit 0
-fi
-
-# Skip fixup/squash commits
-if echo "$COMMIT_MSG" | grep -qE '^(fixup|squash)! '; then
-    exit 0
-fi
-
-# Conventional commit regex
-# Format: type(scope): description
-# Types: feat, fix, chore, docs, refactor, test
-PATTERN='^(feat|fix|chore|docs|refactor|test)(\([a-zA-Z0-9_-]+\))?: .{1,}'
-
-if ! echo "$COMMIT_MSG" | head -1 | grep -qE "$PATTERN"; then
-    echo -e "${RED}=================================================${NC}"
-    echo -e "${RED}ERROR: Commit message doesn't follow conventional format${NC}"
-    echo -e "${RED}=================================================${NC}"
-    echo ""
-    echo "Your message: $(head -1 "$COMMIT_MSG_FILE")"
-    echo ""
-    echo "Expected format:"
-    echo "  <type>(<scope>): <description>"
-    echo ""
-    echo "Types:"
-    echo "  feat     - New feature"
-    echo "  fix      - Bug fix"
-    echo "  chore    - Maintenance task"
-    echo "  docs     - Documentation"
-    echo "  refactor - Code refactoring"
-    echo "  test     - Tests"
-    echo ""
-    echo "Examples:"
-    echo "  feat(auth): add OAuth2 flow support"
-    echo "  fix(api): handle null response from endpoint"
-    echo "  chore(deps): update Helm dependencies"
-    echo "  docs: update README with deployment steps"
-    echo ""
-    exit 1
-fi
-
-# Check description length (warn if too long)
-FIRST_LINE=$(head -1 "$COMMIT_MSG_FILE")
-if [[ ${#FIRST_LINE} -gt 72 ]]; then
-    echo -e "${YELLOW}WARNING: First line is ${#FIRST_LINE} characters (recommended: 72 max)${NC}"
-fi
-
-exit 0

package/.opencode/skills/git-workflow-skill/hooks/pre-commit

@@ -1,38 +0,0 @@
-#!/bin/bash
-# pre-commit hook - Block commits on protected branches
-# Works in both main repo and worktrees.
-# Install: ./scripts/install-hooks.sh
-
-RED='\033[0;31m'
-YELLOW='\033[1;33m'
-NC='\033[0m'
-
-CURRENT_BRANCH=$(git rev-parse --abbrev-ref HEAD)
-
-# Block commits on protected branches
-PROTECTED_BRANCHES="dev main master"
-
-for branch in $PROTECTED_BRANCHES; do
-    if [[ "$CURRENT_BRANCH" == "$branch" ]]; then
-        echo -e "${RED}=================================================${NC}"
-        echo -e "${RED}ERROR: Direct commits to '$branch' are not allowed${NC}"
-        echo -e "${RED}=================================================${NC}"
-        echo ""
-        echo "Create an isolated worktree for your changes:"
-        echo "  ./scripts/start-feature.sh feature <description>"
-        echo ""
-        echo "Or manually:"
-        echo "  git worktree add .worktrees/<desc> -b feature/<desc> origin/dev"
-        echo ""
-        exit 1
-    fi
-done
-
-# Warn if branch doesn't follow naming convention
-if ! echo "$CURRENT_BRANCH" | grep -qE '^(feature|bugfix|hotfix|chore)/'; then
-    echo -e "${YELLOW}WARNING: Branch '$CURRENT_BRANCH' doesn't follow naming convention${NC}"
-    echo -e "${YELLOW}Expected: feature|bugfix|hotfix|chore/<description>${NC}"
-    echo ""
-fi
-
-exit 0

package/.opencode/skills/git-workflow-skill/hooks/prepare-commit-msg

@@ -1,56 +0,0 @@
-#!/bin/bash
-# prepare-commit-msg hook - Add template for conventional commits
-# Install: cp this file to .git/hooks/prepare-commit-msg && chmod +x .git/hooks/prepare-commit-msg
-
-COMMIT_MSG_FILE="$1"
-COMMIT_SOURCE="$2"
-
-# Only add template for new commits (not amend, merge, etc.)
-if [[ -z "$COMMIT_SOURCE" ]]; then
-    # Get current branch to suggest type
-    CURRENT_BRANCH=$(git rev-parse --abbrev-ref HEAD 2>/dev/null)
-    
-    # Extract type from branch name if possible
-    SUGGESTED_TYPE=""
-    if echo "$CURRENT_BRANCH" | grep -qE '^feature/'; then
-        SUGGESTED_TYPE="feat"
-    elif echo "$CURRENT_BRANCH" | grep -qE '^bugfix/'; then
-        SUGGESTED_TYPE="fix"
-    elif echo "$CURRENT_BRANCH" | grep -qE '^hotfix/'; then
-        SUGGESTED_TYPE="fix"
-    elif echo "$CURRENT_BRANCH" | grep -qE '^chore/'; then
-        SUGGESTED_TYPE="chore"
-    fi
-
-    # Only add template if commit message is empty or just comments
-    if ! grep -qv '^#' "$COMMIT_MSG_FILE" 2>/dev/null || [[ ! -s "$COMMIT_MSG_FILE" ]]; then
-        {
-            if [[ -n "$SUGGESTED_TYPE" ]]; then
-                echo "${SUGGESTED_TYPE}(scope): "
-            else
-                echo "type(scope): "
-            fi
-            echo ""
-            echo "# Conventional Commit Format"
-            echo "# --------------------------"
-            echo "# <type>(<scope>): <description>"
-            echo "#"
-            echo "# Types:"
-            echo "#   feat     - New feature"
-            echo "#   fix      - Bug fix"
-            echo "#   chore    - Maintenance"
-            echo "#   docs     - Documentation"
-            echo "#   refactor - Code refactoring"
-            echo "#   test     - Tests"
-            echo "#"
-            echo "# Scope: Component or area (optional)"
-            echo "# Description: Short summary in present tense"
-            echo "#"
-            echo "# Examples:"
-            echo "#   feat(auth): add OAuth2 support"
-            echo "#   fix(api): handle null response"
-        } > "$COMMIT_MSG_FILE"
-    fi
-fi
-
-exit 0