compound-engineering-pi 0.2.3 → 0.2.4

package/README.md

@@ -1,124 +1,76 @@
 # Compound Engineering for Pi
 
-[![Build Status](https://github.com/gvkhosla/compound-engineering-pi/actions/workflows/ci.yml/badge.svg)](https://github.com/gvkhosla/compound-engineering-pi/actions/workflows/ci.yml)
 [![npm](https://img.shields.io/npm/v/compound-engineering-pi)](https://www.npmjs.com/package/compound-engineering-pi)
+[![Build Status](https://github.com/gvkhosla/compound-engineering-pi/actions/workflows/ci.yml/badge.svg)](https://github.com/gvkhosla/compound-engineering-pi/actions/workflows/ci.yml)
 
-Pi-first extension that takes the original Compound Engineering approach and makes it work smoothly in **Pi**.
-
-## 🙌 Credit to the original creators
-
-This project builds directly on the original work by **Every** and **Kieran**.
-
-- Upstream plugin (source of truth): https://github.com/EveryInc/compound-engineering-plugin
-- Compound engineering writeup: https://every.to/chain-of-thought/compound-engineering-how-every-codes-with-agents
-
-If you’re using Claude Code, Codex, OpenCode, or anything non-Pi-focused, use the upstream repo above.
-
----
-
-## Why this exists
-
-The Every plugin is excellent, but Pi has different runtime semantics. This extension adapts the workflow for Pi so it feels native and works out of the box.
-
-### What this Pi extension adds
-
-- Native Pi package install
-- Prebuilt Pi resources (`prompts/`, `skills/`, `extensions/`)
-- Compatibility tool layer for Claude-style workflows
-- Subagent orchestration in Pi (`single`, `parallel`, `chain`)
-- MCP compatibility through **MCPorter** (`mcporter_list`, `mcporter_call`)
-
----
-
-## Install (recommended)
-
-### 1) Install Pi
-
-- Pi project: https://github.com/mariozechner/pi-coding-agent
-- Pi package gallery: https://shittycodingagent.ai/packages
-- Discovery request opened with Pi maintainers: https://github.com/badlogic/pi-mono/issues/1480
+Compound Engineering adapted for [Pi](https://github.com/mariozechner/pi-coding-agent). 47 skills, 9 workflow prompts, and a native compatibility layer, all installable in one command.
 
-### 2) Install MCPorter
+## Install
 
 ```bash
-npm i -g mcporter
-```
-
-### 3) Install this extension in Pi
-
-```bash
-# after npm publish:
 pi install npm:compound-engineering-pi -l
-# fallback (works now):
-pi install git:github.com/gvkhosla/compound-engineering-pi@v0.2.3 -l
 ```
 
-### 4) Reload and test
+Reload your session, then you're ready:
 
 ```bash
 /reload
-pi -p "/workflows-plan Build a todo app"
 ```
 
-Maintainer publish shortcut (first run `npm login` once):
+## Quick start
 
 ```bash
-npm run release:ship
+pi -p "/workflows-plan Build a todo app"
+pi -p "/workflows-work"
+pi -p "/workflows-review"
+pi -p "/workflows-compound"
 ```
 
----
-
-## What you get immediately in Pi
+These follow the core loop: **Plan, Work, Review, Compound**.
 
-- Workflow prompts like `/workflows-plan`, `/workflows-work`, `/workflows-review`, `/workflows-compound`
-- Converted skills and reviewer agents
-- Pi compatibility extension with tools:
-  - `ask_user_question`
-  - `subagent`
-  - `mcporter_list`
-  - `mcporter_call`
+## What's included
 
-This keeps the same compounding loop in Pi:
+### Workflow prompts
 
-**Plan → Work → Review → Compound**
+| Prompt | What it does |
+|--------|-------------|
+| `/workflows-plan` | Create a structured project plan |
+| `/workflows-work` | Execute work from a plan |
+| `/workflows-review` | Multi-agent code review |
+| `/workflows-compound` | Document learnings for future sessions |
+| `/workflows-brainstorm` | Explore ideas before planning |
+| `/deepen-plan` | Add depth to an existing plan |
+| `/test-browser` | Run browser tests on your pages |
+| `/feature-video` | Record a video walkthrough of a feature |
+| `/resolve_todo_parallel` | Resolve all pending todos in parallel |
 
----
+### 47 skills
 
-## How MCP works here
+Code review, research, design, workflow automation, and more. Specialized reviewers for Rails, Python, and TypeScript. Security auditing, architecture analysis, design iteration, git worktree management, image generation, and browser automation.
 
-This extension uses **MCPorter** for MCP interoperability in Pi.
+Run `pi skills` after installing to see the full list.
 
-- No assumption of Claude-native MCP runtime behavior
-- Uses `mcporter` CLI under the hood
-- Supports project/global config and bundled fallback config
+### Pi compatibility extension
 
-MCPorter project: https://github.com/steipete/mcporter
+Four tools that bridge common agent workflows into Pi:
 
----
+- `ask_user_question` - prompt the user for input
+- `subagent` - orchestrate tasks (`single`, `parallel`, `chain`)
+- `mcporter_list` - list available MCP servers
+- `mcporter_call` - call MCP server tools
 
-## Advanced (Pi generation workflow)
+### Optional: MCP support via MCPorter
 
-If you want to regenerate Pi output from the upstream plugin manually:
+For MCP interoperability, install [MCPorter](https://github.com/steipete/mcporter):
 
 ```bash
-bunx compound-engineering-pi install compound-engineering --to pi --pi-home ~/.pi/agent
+npm i -g mcporter
 ```
 
----
-
-## Non-Pi usage (upstream)
-
-For all non-Pi usage and canonical plugin docs, use:
-
-- https://github.com/EveryInc/compound-engineering-plugin
-
-That includes Claude Code-native setup, canonical command/skill design, and upstream issues/roadmap.
+## Credits
 
----
+Built on the [Compound Engineering plugin](https://github.com/EveryInc/compound-engineering-plugin) by [Every](https://every.to) and [Kieran Klaassen](https://github.com/kieranklaassen). Read the [original writeup](https://every.to/chain-of-thought/compound-engineering-how-every-codes-with-agents).
 
-## Learn more
+## License
 
-- Pi guide for this extension: [docs/pi.md](docs/pi.md)
-- Validation report: [VALIDATION.md](VALIDATION.md)
-- npm release checklist: [NPM_PUBLISH.md](NPM_PUBLISH.md)
-- Upstream canonical component docs: https://github.com/EveryInc/compound-engineering-plugin/blob/main/plugins/compound-engineering/README.md
+MIT

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "compound-engineering-pi",
-  "version": "0.2.3",
+  "version": "0.2.4",
   "type": "module",
   "private": false,
   "description": "Pi-first Compound Engineering extension with subagents and MCPorter compatibility.",

package/plugins/compound-engineering/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "compound-engineering",
-  "version": "2.31.1",
+  "version": "2.31.2",
   "description": "AI-powered development tools. 29 agents, 24 commands, 18 skills, 1 MCP server for code review, research, design, and workflow automation.",
   "author": {
     "name": "Kieran Klaassen",

package/plugins/compound-engineering/CHANGELOG.md

@@ -5,6 +5,12 @@ All notable changes to the compound-engineering plugin will be documented in thi
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [2.31.2] - 2026-02-13
+
+### Fixed
+
+- **`resolve-pr-parallel` skill naming compliance** — Updated skill frontmatter name to `resolve-pr-parallel` (hyphen-case) and aligned command references in docs to avoid Pi package validation warnings for invalid skill names.
+
 ## [2.31.1] - 2026-02-09
 
 ### Changed

package/plugins/compound-engineering/README.md

@@ -98,7 +98,7 @@ Core workflow commands use `workflows:` prefix to avoid collisions with built-in
 | `/report-bug` | Report a bug in the plugin |
 | `/reproduce-bug` | Reproduce bugs using logs and console |
 | `/resolve_parallel` | Resolve TODO comments in parallel |
-| `/resolve_pr_parallel` | Resolve PR comments in parallel |
+| `/resolve-pr-parallel` | Resolve PR comments in parallel |
 | `/resolve_todo_parallel` | Resolve todos in parallel |
 | `/triage` | Triage and prioritize issues |
 | `/test-browser` | Run browser tests on PR-affected pages |

package/plugins/compound-engineering/commands/deepen-plan.md

@@ -480,18 +480,20 @@ After writing the enhanced plan, use the **AskUserQuestion tool** to present the
 
 **Options:**
 1. **View diff** - Show what was added/changed
-2. **Run `/technical_review`** - Get feedback from reviewers on enhanced plan
-3. **Start `/workflows:work`** - Begin implementing this enhanced plan
+2. **Use `/technical_review`** - Get feedback from reviewers on enhanced plan
+3. **Use `/workflows:work`** - Begin implementing this enhanced plan
 4. **Deepen further** - Run another round of research on specific sections
 5. **Revert** - Restore original plan (if backup exists)
 
 Based on selection:
 - **View diff** → Run `git diff [plan_path]` or show before/after
-- **`/technical_review`** → Call the /technical_review command with the plan file path
-- **`/workflows:work`** → Call the /workflows:work command with the plan file path
+- **`/technical_review`** → Run `pi --no-session -p "/technical_review [plan_path]"`
+- **`/workflows:work`** → Run `pi --no-session -p "/workflows:work [plan_path]"`
 - **Deepen further** → Ask which sections need more research, then re-run those agents
 - **Revert** → Restore from git or backup
 
+**Important:** Slash commands (like `/workflows:work`) are prompt templates, not shell executables. Never run `/...` directly via bash.
+
 ## Example Enhancement
 
 **Before (from /workflows:plan):**

package/plugins/compound-engineering/commands/test-xcode.md

@@ -328,5 +328,5 @@ mcp__xcodebuildmcp__shutdown_simulator({ simulator_id: "[uuid]" })
 When reviewing PRs that touch iOS code, the `/workflows:review` command can spawn this as a subagent:
 
 ```
-Task general-purpose("Run /xcode-test for scheme [name]. Build, install on simulator, test key screens, check for crashes.")
+Task general-purpose("Invoke `/xcode-test` as a prompt for scheme [name]. Build, install on simulator, test key screens, check for crashes. Never run `/xcode-test` directly via bash.")
 ```

package/plugins/compound-engineering/commands/triage.md

@@ -305,7 +305,7 @@ When done give these options
 ```markdown
 What would you like to do next?
 
-1. run /resolve_todo_parallel to resolve the todos
+1. invoke `/resolve_todo_parallel` as a prompt to resolve the todos
 2. commit the todos
 3. nothing, go chill
 ```

package/plugins/compound-engineering/commands/workflows/brainstorm.md

@@ -86,7 +86,7 @@ Use **AskUserQuestion tool** to present next steps:
 
 **Options:**
 1. **Review and refine** - Improve the document through structured self-review
-2. **Proceed to planning** - Run `/workflows:plan` (will auto-detect this brainstorm)
+2. **Proceed to planning** - Invoke `/workflows:plan` as a prompt (will auto-detect this brainstorm)
 3. **Done for now** - Return later
 
 **If user selects "Review and refine":**
@@ -95,8 +95,8 @@ Load the `document-review` skill and apply it to the brainstorm document.
 
 When document-review returns "Review complete", present next steps:
 
-1. **Move to planning** - Continue to `/workflows:plan` with this document
-2. **Done for now** - Brainstorming complete. To start planning later: `/workflows:plan [document-path]`
+1. **Move to planning** - Continue by invoking `/workflows:plan` with this document
+2. **Done for now** - Brainstorming complete. To start planning later, invoke `/workflows:plan [document-path]`
 
 ## Output Summary
 
@@ -111,7 +111,7 @@ Key decisions:
 - [Decision 1]
 - [Decision 2]
 
-Next: Run `/workflows:plan` when ready to implement.
+Next: Invoke `/workflows:plan` when ready to implement.
 ```
 
 ## Important Guidelines

package/plugins/compound-engineering/commands/workflows/plan.md

@@ -497,24 +497,26 @@ After writing the plan file, use the **AskUserQuestion tool** to present these o
 
 **Options:**
 1. **Open plan in editor** - Open the plan file for review
-2. **Run `/deepen-plan`** - Enhance each section with parallel research agents (best practices, performance, UI)
-3. **Run `/technical_review`** - Technical feedback from code-focused reviewers (DHH, Kieran, Simplicity)
+2. **Use `/deepen-plan`** - Enhance each section with parallel research agents (best practices, performance, UI)
+3. **Use `/technical_review`** - Technical feedback from code-focused reviewers (DHH, Kieran, Simplicity)
 4. **Review and refine** - Improve the document through structured self-review
-5. **Start `/workflows:work`** - Begin implementing this plan locally
-6. **Start `/workflows:work` on remote** - Begin implementing in Claude Code on the web (use `&` to run in background)
+5. **Use `/workflows:work`** - Begin implementing this plan locally
+6. **Use `/workflows:work` on remote** - Begin implementing in Claude Code on the web (use `&` to run in background)
 7. **Create Issue** - Create issue in project tracker (GitHub/Linear)
 
 Based on selection:
 - **Open plan in editor** → Run `open docs/plans/<plan_filename>.md` to open the file in the user's default editor
-- **`/deepen-plan`** → Call the /deepen-plan command with the plan file path to enhance with research
-- **`/technical_review`** → Call the /technical_review command with the plan file path
+- **`/deepen-plan`** → Run `pi --no-session -p "/deepen-plan docs/plans/<plan_filename>.md"` to enhance with research
+- **`/technical_review`** → Run `pi --no-session -p "/technical_review docs/plans/<plan_filename>.md"`
 - **Review and refine** → Load `document-review` skill.
-- **`/workflows:work`** → Call the /workflows:work command with the plan file path
-- **`/workflows:work` on remote** → Run `/workflows:work docs/plans/<plan_filename>.md &` to start work in background for Claude Code web
+- **`/workflows:work`** → Run `pi --no-session -p "/workflows:work docs/plans/<plan_filename>.md"`
+- **`/workflows:work` on remote** → Run `pi --no-session -p "/workflows:work docs/plans/<plan_filename>.md" &` to start work in background
 - **Create Issue** → See "Issue Creation" section below
 - **Other** (automatically provided) → Accept free text for rework or specific changes
 
-**Note:** If running `/workflows:plan` with ultrathink enabled, automatically run `/deepen-plan` after plan creation for maximum depth and grounding.
+**Important:** Slash commands (like `/deepen-plan`) are prompt templates, not shell executables. Never run `/...` directly via bash.
+
+**Note:** If running `/workflows:plan` with ultrathink enabled, automatically run `/deepen-plan` after plan creation by invoking `pi --no-session -p "/deepen-plan docs/plans/<plan_filename>.md"`.
 
 Loop back to options after Simplify or Other changes until user selects `/workflows:work` or `/technical_review`.
 

package/plugins/compound-engineering/commands/workflows/review.md

@@ -485,7 +485,7 @@ After presenting the Summary Report, offer appropriate testing based on project
 Spawn a subagent to run browser tests (preserves main context):
 
 ```
-Task general-purpose("Run /test-browser for PR #[number]. Test all affected pages, check for console errors, handle failures by creating todos and fixing.")
+Task general-purpose("Invoke `/test-browser` as a prompt for PR #[number]. Test all affected pages, check for console errors, handle failures by creating todos and fixing. Never run `/test-browser` directly via bash.")
 ```
 
 The subagent will:
@@ -504,7 +504,7 @@ The subagent will:
 Spawn a subagent to run Xcode tests (preserves main context):
 
 ```
-Task general-purpose("Run /xcode-test for scheme [name]. Build for simulator, install, launch, take screenshots, check for crashes.")
+Task general-purpose("Invoke `/xcode-test` as a prompt for scheme [name]. Build for simulator, install, launch, take screenshots, check for crashes. Never run `/xcode-test` directly via bash.")
 ```
 
 The subagent will:

package/plugins/compound-engineering/skills/file-todos/SKILL.md

@@ -186,7 +186,7 @@ Work logs serve as:
 | Trigger | Flow | Tool |
 |---------|------|------|
 | Code review | `/workflows:review` → Findings → `/triage` → Todos | Review agent + skill |
-| PR comments | `/resolve_pr_parallel` → Individual fixes → Todos | gh CLI + skill |
+| PR comments | `/resolve-pr-parallel` → Individual fixes → Todos | gh CLI + skill |
 | Code TODOs | `/resolve_todo_parallel` → Fixes + Complex todos | Agent + skill |
 | Planning | Brainstorm → Create todo → Work → Complete | Skill |
 | Feedback | Discussion → Create todo → Triage → Work | Skill + slash |

package/plugins/compound-engineering/skills/resolve-pr-parallel/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: resolve_pr_parallel
+name: resolve-pr-parallel
 description: Resolve all PR comments using parallel processing. Use when addressing PR review feedback, resolving review threads, or batch-fixing PR comments.
 argument-hint: "[optional: PR number or current PR]"
 disable-model-invocation: true

package/prompts/deepen-plan.md

@@ -479,18 +479,20 @@ After writing the enhanced plan, use the **ask_user_question tool** to present t
 
 **Options:**
 1. **View diff** - Show what was added/changed
-2. **Run `/technical_review`** - Get feedback from reviewers on enhanced plan
-3. **Start `/workflows-work`** - Begin implementing this enhanced plan
+2. **Use `/technical_review`** - Get feedback from reviewers on enhanced plan
+3. **Use `/workflows-work`** - Begin implementing this enhanced plan
 4. **Deepen further** - Run another round of research on specific sections
 5. **Revert** - Restore original plan (if backup exists)
 
 Based on selection:
 - **View diff** → Run `git diff [plan_path]` or show before/after
-- **`/technical_review`** → Call the /technical_review command with the plan file path
-- **`/workflows-work`** → Call the /workflows-work command with the plan file path
+- **`/technical_review`** → Run `pi --no-session -p "/technical_review [plan_path]"`
+- **`/workflows-work`** → Run `pi --no-session -p "/workflows-work [plan_path]"`
 - **Deepen further** → Ask which sections need more research, then re-run those agents
 - **Revert** → Restore from git or backup
 
+**Important:** Slash commands (like `/workflows-work`) are Pi prompt templates, not shell executables. Never run `/...` directly via bash.
+
 ## Example Enhancement
 
 **Before (from /workflows-plan):**

package/prompts/workflows-brainstorm.md

@@ -85,7 +85,7 @@ Use **ask_user_question tool** to present next steps:
 
 **Options:**
 1. **Review and refine** - Improve the document through structured self-review
-2. **Proceed to planning** - Run `/workflows-plan` (will auto-detect this brainstorm)
+2. **Proceed to planning** - Invoke `/workflows-plan` as a Pi prompt (will auto-detect this brainstorm)
 3. **Done for now** - Return later
 
 **If user selects "Review and refine":**
@@ -94,8 +94,8 @@ Load the `document-review` skill and apply it to the brainstorm document.
 
 When document-review returns "Review complete", present next steps:
 
-1. **Move to planning** - Continue to `/workflows-plan` with this document
-2. **Done for now** - Brainstorming complete. To start planning later: `/workflows-plan [document-path]`
+1. **Move to planning** - Continue by invoking `/workflows-plan` with this document
+2. **Done for now** - Brainstorming complete. To start planning later, invoke `/workflows-plan [document-path]` as a Pi prompt
 
 ## Output Summary
 
@@ -110,7 +110,7 @@ Key decisions:
 - [Decision 1]
 - [Decision 2]
 
-Next: Run `/workflows-plan` when ready to implement.
+Next: Invoke `/workflows-plan` as a Pi prompt when ready to implement.
 ```
 
 ## Important Guidelines

package/prompts/workflows-plan.md

@@ -496,24 +496,26 @@ After writing the plan file, use the **ask_user_question tool** to present these
 
 **Options:**
 1. **Open plan in editor** - Open the plan file for review
-2. **Run `/deepen-plan`** - Enhance each section with parallel research agents (best practices, performance, UI)
-3. **Run `/technical_review`** - Technical feedback from code-focused reviewers (DHH, Kieran, Simplicity)
+2. **Use `/deepen-plan`** - Enhance each section with parallel research agents (best practices, performance, UI)
+3. **Use `/technical_review`** - Technical feedback from code-focused reviewers (DHH, Kieran, Simplicity)
 4. **Review and refine** - Improve the document through structured self-review
-5. **Start `/workflows-work`** - Begin implementing this plan locally
-6. **Start `/workflows-work` on remote** - Begin implementing in Claude Code on the web (use `&` to run in background)
+5. **Use `/workflows-work`** - Begin implementing this plan locally
+6. **Use `/workflows-work` on remote** - Begin implementing in Claude Code on the web (use `&` to run in background)
 7. **Create Issue** - Create issue in project tracker (GitHub/Linear)
 
 Based on selection:
 - **Open plan in editor** → Run `open docs/plans/<plan_filename>.md` to open the file in the user's default editor
-- **`/deepen-plan`** → Call the /deepen-plan command with the plan file path to enhance with research
-- **`/technical_review`** → Call the /technical_review command with the plan file path
+- **`/deepen-plan`** → Run `pi --no-session -p "/deepen-plan docs/plans/<plan_filename>.md"` to enhance with research
+- **`/technical_review`** → Run `pi --no-session -p "/technical_review docs/plans/<plan_filename>.md"`
 - **Review and refine** → Load `document-review` skill.
-- **`/workflows-work`** → Call the /workflows-work command with the plan file path
-- **`/workflows-work` on remote** → Run `/workflows-work docs/plans/<plan_filename>.md &` to start work in background for Claude Code web
+- **`/workflows-work`** → Run `pi --no-session -p "/workflows-work docs/plans/<plan_filename>.md"`
+- **`/workflows-work` on remote** → Run `pi --no-session -p "/workflows-work docs/plans/<plan_filename>.md" &` to start work in background
 - **Create Issue** → See "Issue Creation" section below
 - **Other** (automatically provided) → Accept free text for rework or specific changes
 
-**Note:** If running `/workflows-plan` with ultrathink enabled, automatically run `/deepen-plan` after plan creation for maximum depth and grounding.
+**Important:** Slash commands (like `/deepen-plan`) are Pi prompt templates, not shell executables. Never run `/...` directly via bash.
+
+**Note:** If running `/workflows-plan` with ultrathink enabled, automatically run `/deepen-plan` after plan creation by invoking `pi --no-session -p "/deepen-plan docs/plans/<plan_filename>.md"`.
 
 Loop back to options after Simplify or Other changes until user selects `/workflows-work` or `/technical_review`.
 

package/prompts/workflows-review.md

@@ -484,7 +484,7 @@ After presenting the Summary Report, offer appropriate testing based on project
 Spawn a subagent to run browser tests (preserves main context):
 
 ```
-Run subagent with agent="general-purpose" and task=""Run /test-browser for PR #[number]. Test all affected pages, check for console errors, handle failures by creating todos and fixing."".
+Run subagent with agent="general-purpose" and task=""Invoke `/test-browser` as a Pi prompt for PR #[number]. Test all affected pages, check for console errors, handle failures by creating todos and fixing. Never run `/test-browser` directly via bash."".
 ```
 
 The subagent will:
@@ -503,7 +503,7 @@ The subagent will:
 Spawn a subagent to run Xcode tests (preserves main context):
 
 ```
-Run subagent with agent="general-purpose" and task=""Run /xcode-test for scheme [name]. Build for simulator, install, launch, take screenshots, check for crashes."".
+Run subagent with agent="general-purpose" and task=""Invoke `/xcode-test` as a Pi prompt for scheme [name]. Build for simulator, install, launch, take screenshots, check for crashes. Never run `/xcode-test` directly via bash."".
 ```
 
 The subagent will:

package/skills/file-todos/SKILL.md

@@ -186,7 +186,7 @@ Work logs serve as:
 | Trigger | Flow | Tool |
 |---------|------|------|
 | Code review | `/workflows:review` → Findings → `/triage` → Todos | Review agent + skill |
-| PR comments | `/resolve_pr_parallel` → Individual fixes → Todos | gh CLI + skill |
+| PR comments | `/resolve-pr-parallel` → Individual fixes → Todos | gh CLI + skill |
 | Code TODOs | `/resolve_todo_parallel` → Fixes + Complex todos | Agent + skill |
 | Planning | Brainstorm → Create todo → Work → Complete | Skill |
 | Feedback | Discussion → Create todo → Triage → Work | Skill + slash |

package/skills/{resolve_pr_parallel → resolve-pr-parallel}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: resolve_pr_parallel
+name: resolve-pr-parallel
 description: Resolve all PR comments using parallel processing. Use when addressing PR review feedback, resolving review threads, or batch-fixing PR comments.
 argument-hint: "[optional: PR number or current PR]"
 disable-model-invocation: true

package/src/converters/claude-to-pi.ts

@@ -127,6 +127,36 @@ function transformContentForPi(body: string): string {
     return `/${normalizeName(withoutPrefix)}`
   })
 
+  result = rewriteSlashCommandExecutionForPi(result)
+
+  return result
+}
+
+function rewriteSlashCommandExecutionForPi(body: string): string {
+  let result = body
+
+  result = result.replace(/Call the \/([a-z][a-z0-9_-]*) command/gi, (_match, commandName: string) => {
+    return `Invoke \`/${commandName}\` as a Pi prompt (never as a direct bash command)`
+  })
+
+  result = result.replace(/Run `\/([a-z][a-z0-9_-]*)([^`]*)`/gi, (_match, commandName: string, rawArgs: string) => {
+    let args = (rawArgs ?? "").trim()
+    let backgroundSuffix = ""
+
+    if (args.endsWith("&")) {
+      args = args.slice(0, -1).trimEnd()
+      backgroundSuffix = " &"
+    }
+
+    const prompt = args.length > 0 ? `/${commandName} ${args}` : `/${commandName}`
+    const escapedPrompt = prompt.replace(/"/g, '\\"')
+    return `Run \`pi --no-session -p "${escapedPrompt}"${backgroundSuffix}\``
+  })
+
+  if (result !== body && !result.includes("Slash commands are Pi prompt templates")) {
+    result += "\n\n**Important:** Slash commands are Pi prompt templates, not shell executables. Never run `/...` directly via bash."
+  }
+
   return result
 }