@hailer/mcp 1.1.12 → 1.1.13

package/.opencode/commands/prd.md

@@ -1,131 +0,0 @@
----
-description: Create Product Requirements Document (PRD)
----
-
-# Create Product Requirements Document (PRD)
-
-Create a structured PRD for a feature.
-
-## Arguments
-- `$ARGUMENTS` - Feature name or description
-
-## Process
-
-### Step 1: Gather Information
-If feature name is vague, ask clarifying questions:
-1. What problem does this solve?
-2. Who is the target user?
-3. What are the key capabilities needed?
-
-### Step 2: Generate PRD
-
-Create `docs/prd-[feature-name].md`:
-
-```markdown
-# PRD: [Feature Name]
-
-**Created:** YYYY-MM-DD
-**Status:** Draft
-
----
-
-## Executive Summary
-[1-2 paragraph overview of the feature and its value]
-
----
-
-## Problem Statement
-**Current State:** [How things work now]
-**Pain Points:**
-- [Pain point 1]
-- [Pain point 2]
-
-**Desired Outcome:** [What success looks like]
-
----
-
-## User Stories
-
-### Primary User: [Role]
-- As a [role], I want [capability] so that [benefit]
-
-### Secondary User: [Role]
-- As a [role], I want [capability] so that [benefit]
-
----
-
-## Acceptance Criteria
-
-### Must Have (P0)
-- [ ] Criterion 1
-- [ ] Criterion 2
-
-### Should Have (P1)
-- [ ] Criterion 1
-
-### Nice to Have (P2)
-- [ ] Criterion 1
-
----
-
-## Technical Constraints
-
-### Performance
-- [Response time requirements]
-
-### Security
-- [Auth/data protection needs]
-
-### Compatibility
-- [Browser/device/integration requirements]
-
----
-
-## Out of Scope
-- [Item 1]
-- [Item 2]
-
----
-
-## Success Metrics
-| Metric | Target | Measurement |
-|--------|--------|-------------|
-| [Metric] | [Target] | [How measured] |
-
----
-
-## Open Questions
-- [ ] Question 1
-- [ ] Question 2
-
----
-
-## Dependencies
-- [Dependency 1]
-```
-
-### Step 3: Update DEVELOPMENT.md (Required)
-
-**Always update DEVELOPMENT.md after creating a PRD:**
-
-1. If DEVELOPMENT.md doesn't exist, create it first (use `/project-status`)
-2. Add the new PRD to the Roadmap table:
-
-```markdown
-| [Feature Name] | [docs/prd-feature-name.md](docs/prd-feature-name.md) | Planned |
-```
-
-3. Update "Current Status" section if this is the active work
-
-### Step 4: Confirm
-
-Report to user:
-```
-Created: docs/prd-[feature-name].md
-Updated: DEVELOPMENT.md roadmap
-
-Next steps:
-- Review PRD requirements
-- When ready to implement, `/yolo` works well for PRD-based tasks
-  since the requirements are already clearly defined
-```

package/.opencode/commands/project-status.md

@@ -1,82 +0,0 @@
----
-description: Review and update DEVELOPMENT.md with current project state
----
-
-# Sync Project Status
-
-Review and update DEVELOPMENT.md with current project state.
-
-## Process
-
-### Step 1: Read Current State
-
-1. Read `DEVELOPMENT.md` if it exists
-2. Read `SESSION-HANDOFF.md` if it exists
-3. Check `docs/prd-*.md` files for PRD statuses
-4. Check git status for uncommitted changes
-5. Review current TodoWrite items
-
-### Step 2: Sync PRD Statuses
-
-For each PRD in `docs/prd-*.md`:
-- Check if implementation is complete (all checkboxes checked)
-- Ensure PRD Status field matches actual state (Draft/In Progress/Complete)
-- Ensure DEVELOPMENT.md Roadmap matches PRD status
-
-### Step 3: Update DEVELOPMENT.md
-
-Update or create DEVELOPMENT.md following this structure:
-
-```markdown
-# [Project Name]
-
-**Stack:** [tech stack]
-**Entry:** [main entry point]
-**Run:** [dev command]
-
----
-
-## What This Project Does
-[2-3 sentences]
-
----
-
-## Architecture
-[Key directories and patterns]
-
----
-
-## Roadmap
-
-| Feature | PRD | Status |
-|---------|-----|--------|
-| [Feature] | [docs/prd-*.md] | [Status] |
-
----
-
-## Current Status
-**Active:** [Current work]
-**Blocked:** [Blockers or "None"]
-
----
-
-## Known Issues
-- [ ] Issue (BUG/HIGH)
-- [ ] Note (NOTE)
-
----
-
-## Technical Decisions
-| Decision | Rationale | Date |
-|----------|-----------|------|
-```
-
-### Step 4: Report Changes
-
-Tell user what was updated and suggest next actions.
-
-## Notes
-
-- This is a local-only sync (no external services)
-- Updates are based on git, todos, and session context
-- Use `/handoff` before ending a session

package/.opencode/commands/publish.md

@@ -1,138 +0,0 @@
----
-description: Publish to Marketplace
----
-
-# Publish to Marketplace
-
-Publish plugins from working copy (`.claude/`) to the Hailer Marketplace.
-
-## Usage
-
-```
-/publish <plugin-name>           # Publish single plugin
-/publish all                     # Publish all out-of-sync plugins
-```
-
-## Arguments
-
-- `$ARGUMENTS` contains the publish target
-
-## Paths
-
-- **Working copy:** `.claude/` (source of truth)
-- **Marketplace:** `Hailer-Marketplace/` (git repo)
-- **Individual plugins:** `Hailer-Marketplace/{plugin-name}/`
-
----
-
-## Mode: <plugin-name>
-
-Publishes a single plugin to the marketplace.
-
-### Steps
-
-1. **Find the plugin** in `.claude/`:
-   ```
-   .claude/agents/agent-{name}*.md
-   .claude/skills/{name}*/SKILL.md
-   .claude/hooks/{name}*.cjs
-   ```
-
-2. **Detect plugin type** (agent, skill, hook)
-
-3. **Ask for version and changelog**
-   - Check existing version in `Hailer-Marketplace/{plugin-name}/.claude-plugin/plugin.json`
-   - Suggest patch bump
-
-4. **Invoke Publisher Agent:**
-   ```json
-   {
-     "task": "publish_plugin",
-     "plugin": {
-       "name": "{name}",
-       "type": "agent|skill|hook",
-       "version": "{user provided}",
-       "changelog_message": "{user provided}",
-       "content": "{file content}"
-     }
-   }
-   ```
-
-5. **Invoke Reviewer Agent** (after PR created):
-   ```json
-   {
-     "task": "review_pr",
-     "pr_number": "{from publisher result}"
-   }
-   ```
-
-6. **Report result** - PR merged and tags created, or changes requested
-
----
-
-## Mode: all
-
-Publishes all out-of-sync plugins in a single PR.
-
-### Steps
-
-1. **Invoke Publisher Agent** with version_check to find mismatches
-2. **Show results** to user, ask to confirm
-3. **Invoke Publisher Agent** with publish_all
-4. **Invoke Reviewer Agent**
-5. **Report result**
-
----
-
-## Publisher Agent Protocol
-
-Returns:
-```json
-{
-  "status": "pr_created",
-  "result": {
-    "pr_number": 123,
-    "pr_url": "https://github.com/...",
-    "plugins_updated": ["agent-kenji-data-reader"],
-    "versions": { "agent-kenji-data-reader": "1.0.2" }
-  },
-  "summary": "Created PR #123"
-}
-```
-
-## Reviewer Agent Protocol
-
-After publisher creates PR, the reviewer:
-- Validates plugin.json schemas
-- Checks version increments
-- Scans for security issues
-- If ALL checks pass: auto-merges PR, creates git tags
-- If ANY check fails: requests changes
-
----
-
-## Examples
-
-```
-/publish kenji
-→ Finds .claude/agents/agent-kenji-data-reader.md
-→ Syncs to Hailer-Marketplace/agent-kenji-data-reader/
-→ Creates PR, reviewer validates and merges
-→ Tag: agent-kenji-data-reader@1.0.2
-
-/publish all
-→ Checks manifest vs marketplace for mismatches
-→ Single PR with all out-of-sync plugins
-→ Multiple tags created
-```
-
----
-
-## Error Handling
-
-| Error | Action |
-|-------|--------|
-| Plugin not found | List available plugins in `.claude/` |
-| Version conflict | Show existing version, ask for higher |
-| Reviewer rejects | Show failures, user must fix and re-run |
-| Git push fails | Show error, suggest manual intervention |

package/.opencode/commands/recap.md

@@ -1,69 +0,0 @@
----
-description: Resume session - load handoff, project status, and learnings into context
-argument-hint: [optional focus area]
----
-# Resume Session
-
-Load project context at session start or after context reset. This is the counterpart to `/handoff`.
-
-## Instructions
-
-Read ALL of the following files and present a consolidated briefing. Do NOT delegate these reads - read them directly as the orchestrator.
-
-### Step 1: Read Context Files
-
-Read these files (skip any that don't exist):
-
-1. **SESSION-HANDOFF.md** - What happened last session, pending tasks, next steps
-2. **DEVELOPMENT.md** - Project status, roadmap, known issues
-3. **inbox.md** - Pending learnings from other projects (hub repo only)
-
-### Step 2: Recreate Pending Tasks
-
-If SESSION-HANDOFF.md has a "Pending Tasks" section with items:
-- Use `TaskCreate` for each pending task
-- Set descriptions from the handoff context
-
-### Step 3: Check Git State
-
-Run `git status` and `git log --oneline -5` to see:
-- Any uncommitted changes from last session
-- Recent commit history for context
-
-### Step 4: Present Briefing
-
-Format output as:
-
-```
-## Session Resumed
-
-### Last Session
-[Summary from SESSION-HANDOFF.md - what was done, key decisions]
-
-### Current State
-- **Branch:** [branch name]
-- **Uncommitted changes:** [yes/no, summary if yes]
-- **Active work:** [from DEVELOPMENT.md current status]
-
-### Pending Tasks
-[Recreated tasks, or "None"]
-
-### Next Steps
-[From handoff next steps]
-
-### Learnings Inbox
-[Count of pending items, or "Empty"]
-```
-
-### Step 5: Focus (if argument provided)
-
-If user provided `$ARGUMENTS`:
-- Filter the briefing to focus on that area
-- Read relevant PRD files (docs/prd-*.md) if the focus matches a feature
-
-## Notes
-
-- This command reads files DIRECTLY (no delegation) for speed
-- Complements `/handoff` which writes these same files
-- Run at session start or after `/clear` to restore context
-- Updates SESSION-HANDOFF.md: removes "completed" items, keeps active context

package/.opencode/commands/restore.md

@@ -1,64 +0,0 @@
----
-description: Restore to a previous save point
-allowed-tools: Bash
----
-Help user restore to a previous git save point.
-
-**Step 1: Check git exists**
-- If no `.git` folder, tell user: "No git history. Run /git-init first."
-
-**Step 2: Show recent saves**
-Run: `git log --oneline -10`
-
-Present as a numbered list:
-```
-Recent saves:
-
-  [1] a1b2c3d - 2 hours ago - "Fix login button"
-  [2] e4f5g6h - Yesterday - "Add user form"
-  [3] i7j8k9l - 2 days ago - "Before YOLO: Add dashboard"
-  ...
-
-Which save do you want to restore to? (enter number)
-```
-
-Use `git log --oneline --format="%h - %ar - %s" -10` for nice format.
-
-**Step 3: Confirm with user**
-After user picks a number, show:
-```
-You selected: "Before YOLO: Add dashboard"
-
-This will:
-- Save your current state first (safety backup)
-- Restore all files to that point
-
-Proceed? (yes/no)
-```
-
-**Step 4: Safe restore process**
-1. First, save current state:
-   ```bash
-   git add -A
-   git commit -m "Before restore - safety backup" --allow-empty
-   ```
-
-2. Then restore using revert or reset:
-   ```bash
-   git checkout <commit-id> -- .
-   git add -A
-   git commit -m "Restored to: <original message>"
-   ```
-
-**Step 5: Confirm**
-```
-Done! Restored to "Before YOLO: Add dashboard"
-
-Your previous state was saved as "Before restore - safety backup"
-If you want to undo, run /restore again and pick that save.
-```
-
-**Important:**
-- Always create safety backup first
-- Use checkout to restore files (not hard reset)
-- Keep all history intact

package/.opencode/commands/review-squad.md

@@ -1,152 +0,0 @@
----
-description: Run parallel code review with Svetlana, Lars, and Tanya
-argument-hint: [files, directory, or blank for recent changes]
-allowed-tools: Task, Bash
----
-# Review Squad
-
-Parallel code review with iterative fix loop: find issues, auto-fix criticals, re-review.
-
-**Agents:**
-- **Svetlana** - Code review (bugs, security, best practices)
-- **Lars** - LSP inspection (dead code, unused imports, type errors)
-- **Tanya** - Test runner (run tests, verify builds)
-- **Simple Writer** - Auto-fix critical issues (only in loop iterations)
-
-**Target:** $ARGUMENTS (if blank, use recent git changes)
-
-**Changed files:** Run `git diff --name-only HEAD` via Bash to get recent changes. Use the output as TARGET if no specific files are given.
-
-## Protocol
-
-### Step 1: Determine Target
-
-If `$ARGUMENTS` has specific files/directories, use those.
-Otherwise, use the changed files list above as the target.
-
-If `$ARGUMENTS` contains `--skip-tests`:
-- Remove flag from target
-- Skip Tanya in Step 2
-
-If `$ARGUMENTS` contains `--skip-lsp`:
-- Remove flag from target
-- Skip Lars in Step 2
-
-If `$ARGUMENTS` contains `--no-fix`:
-- Skip the fix loop entirely (old behavior - report only)
-
-### Step 2: Launch Review Agents in Parallel
-
-**Set:** `iteration = 1`
-
-If `--bg` is present, add `run_in_background: true` to all agent calls below and tell the user they'll be notified when complete.
-
-Spawn all review agents simultaneously using multiple Task tool calls in a single message:
-
-**Svetlana:**
-```
-Task(subagent_type="agent-svetlana-code-review", prompt="Review these files for bugs, security issues, and best practices: [TARGET]. [IF iteration > 1: Previous issues that were auto-fixed by Simple Writer: [LIST FIXES APPLIED]. Verify these are resolved and check for any NEW issues introduced by the fixes.] Return verdict (APPROVE/REQUEST_CHANGES), critical count, warning count, and details. Categorize each issue as CRITICAL or WARNING.")
-```
-
-**Lars:**
-```
-Task(subagent_type="agent-lars-code-inspector", prompt="Inspect these files for dead code, unused imports, and type errors: [TARGET]. [IF iteration > 1: Previous issues that were auto-fixed: [LIST FIXES APPLIED]. Verify these are resolved.] Use LSP if available, fall back to tsc/eslint.")
-```
-
-**Tanya:**
-```
-Task(subagent_type="agent-tanya-test-runner", prompt="Run tests associated with these files: [TARGET]. Include build verification. Report pass/fail counts.")
-```
-
-### Step 3: Evaluate Results - Fix or Report
-
-Collect results from all agents. Check for **auto-fixable issues**:
-
-**Auto-fixable** (Simple Writer can handle):
-- Unused imports
-- Dead code removal
-- Missing null checks
-- Simple type errors
-- Test failures with obvious fixes
-
-**NOT auto-fixable** (escalate to user):
-- Architecture concerns
-- Design pattern issues
-- Complex logic bugs
-- Security issues requiring design changes
-- Performance concerns
-
-#### If critical auto-fixable issues found AND `--no-fix` not set AND `iteration < 3`:
-
-Spawn Simple Writer with the specific issues:
-
-```
-Task(subagent_type="agent-simple-writer", prompt="Fix these critical issues found by the review squad:\n\n[LIST AUTO-FIXABLE ISSUES WITH FILE PATHS AND LINE NUMBERS]\n\nFrom Svetlana: [critical issues]\nFrom Lars: [dead code, unused imports]\nFrom Tanya: [test failures if obvious fix]\n\nApply minimal fixes only. Do not refactor or improve beyond the reported issues.")
-```
-
-After Simple Writer finishes:
-- Increment `iteration`
-- Go back to **Step 2** to re-review the fixed code
-- On re-review, include what Simple Writer fixed so agents can verify fixes and check for regressions
-
-#### If no auto-fixable criticals OR `--no-fix` set OR `iteration >= 3`:
-
-Proceed to Step 4 (report).
-
-### Step 4: Report
-
-```markdown
-## Review Squad Report
-
-### Loop Summary
-- Review iterations: [count]
-- Auto-fixes applied: [yes/no, count of issues fixed]
-
-### Svetlana (Code Review)
-**Verdict:** APPROVE / REQUEST_CHANGES
-- Critical: X (Y auto-fixed)
-- Warnings: X
-- [Remaining issues grouped by file]
-
-### Lars (LSP Inspector)
-- Dead code: X items (Y auto-removed)
-- Unused imports: X (Y auto-removed)
-- Type errors: X
-- [Remaining issues grouped by file]
-
-### Tanya (Tests)
-- Tests: X passed, X failed
-- Build: Pass/Fail
-- [Failed test details if any]
-
-### Auto-Fixes Applied
-[If Simple Writer made changes:]
-- [File]: [what was fixed]
-- [File]: [what was fixed]
-
-### Remaining Issues (Manual Action Needed)
-[Issues that couldn't be auto-fixed]
-
-### Summary
-[1-2 sentence overall assessment]
-```
-
-If any agent fails or times out, report partial results from successful agents, note which failed, and offer to re-run individually.
-
-## Options
-
-| Flag | Effect |
-|------|--------|
-| `--skip-tests` | Skip Tanya (faster, no test execution) |
-| `--skip-lsp` | Skip Lars (when LSP not configured) |
-| `--no-fix` | Report only, no auto-fix loop (old behavior) |
-| `--bg` | Run all agents in background mode |
-
-If `--bg` is present, launch all agents with `run_in_background: true` and tell the user they'll be notified when complete.
-
-## Notes
-
-- Max 2 re-review iterations (3 total reviews) to prevent infinite loops
-- Simple Writer only fixes clear-cut issues - ambiguous problems go to user
-- Each re-review is a full parallel sweep, not just Svetlana
-- The `--no-fix` flag gives the old report-only behavior

package/.opencode/commands/save.md

@@ -1,24 +0,0 @@
----
-description: Save current changes to git (local commit)
-argument-hint: <description of changes>
-allowed-tools: Bash
----
-Save current work to git (local only, not pushed):
-
-1. Check if git is initialized. If not, ask user if they want to initialize it first.
-
-2. Run `git status` to see what changed
-
-3. Stage all changes EXCEPT:
-   - .env.local
-   - .env
-   - node_modules/
-   - .DS_Store
-
-4. Create commit with message:
-   - If argument provided: use "$ARGUMENTS"
-   - If no argument: generate brief message based on changed files
-
-5. Show confirmation: "Saved locally. Use `git push` when ready to upload to GitLab."
-
-Keep it simple - this is for local saves only.

package/.opencode/commands/stats.md

@@ -1,19 +0,0 @@
----
-description: Show agent usage statistics
----
-
-# Agent Usage Stats
-
-Run the usage logger to show agent statistics.
-
-```bash
-node .claude/hooks/agent-usage-logger.cjs --stats
-```
-
-**Options:**
-- `/stats` - Show all-time stats across all projects
-- `/stats recent` - Show last 20 agent calls
-- `/stats recent 50` - Show last 50 agent calls
-- `/stats clear` - Clear the usage log
-
-**Note:** Stats are collected from `$HAILER_INBOX_DIR/usage.jsonl` which aggregates usage across all Hailer projects.