Port Feature

Always obey `.docs/guides/mcp-tools.md`. Read it now if not already in context. Run `/primer` first if you have not already this session.

Port Feature

Assess a specific feature or area of functionality in an external (source) project and produce a concrete, step-by-step plan for porting it into the current project. The plan accounts for differences in architecture, tech stack, conventions, and dependencies between the two codebases.

---

Arguments: $ARGUMENTS

---

Phase 0: Parse Arguments & Validate

1. Parse `$ARGUMENTS`:

• The first token that is an absolute path (starts with /) is the <source-path>
• Everything after the path is the <functionality> description
• Example: /home/user/other-app authentication with JWT → source path = /home/user/other-app, functionality = authentication with JWT

2. Validate:

• If <source-path> is missing or not absolute: stop and ask the user for the absolute path to the source project
• If <functionality> is empty: stop and ask the user what functionality to port
• Confirm the source directory exists: use mcp__serena__find_file or Bash ls <source-path> to verify it is accessible

3. Clarifying questions (ask before proceeding if ambiguous):

• Should the port plan produce a task file in wiki/work/tasks/, a report in .docs/port/, or both?
• Is there a preferred output level of detail (high-level steps vs. file-by-file implementation plan)?
• Are there any known constraints (e.g., "must stay in TypeScript", "no new dependencies")?

---

Phase 1: Explore the Source Project

Note: Serena is scoped to the current project. Use Read and Bash to explore the source project at <source-path>. Do NOT use Serena tools on the source project's files.

1a. Project Overview

• Read <source-path>/README.md (if present) for declared purpose, setup, and architecture
• Read <source-path>/package.json / <source-path>/pyproject.toml / <source-path>/go.mod / <source-path>/Cargo.toml (whichever is present) for declared dependencies, scripts, and tech stack
• Run ls <source-path> and ls <source-path>/src (or equivalent source root) to understand the top-level structure
• Check for CLAUDE.md, ARCHITECTURE.md, docs/, or similar for documented design decisions

1b. Locate the Target Functionality

Search for the specific functionality described in <functionality>:

1. Identify candidate directories and files — look for modules, routes, services, components, or packages whose names relate to the functionality

• Use Bash: find <source-path> -type f -name "*.ts" | xargs grep -l "<keyword>" 2>/dev/null | head -20 (adapt extension to the detected language)
• Navigate into the most relevant directories with ls

2. Read key implementation files — read the entry points, main logic files, interfaces/types, and any configuration for the target feature 3. Trace data flow — identify:

• Entry points (API routes, CLI commands, event handlers, exported functions)
• Core logic (services, business rules, transformations)
• Data models / schemas / types
• External dependencies (third-party libraries, APIs, databases, queues)
• Configuration (env vars, config files, feature flags)

4. Read test files — test files often give the clearest picture of intended behaviour; find and read relevant tests

1c. Build a Source Feature Inventory

Produce an internal summary documenting:

SOURCE FEATURE: <feature name>
Language / Runtime: <language + version if visible>
Frameworks / Libraries used by this feature:
  - <lib>: <what role it plays>
Entry points:
  - <file:line> — <description>
Core files:
  - <file> — <responsibility>
Data models / types:
  - <name> — <shape and purpose>
External dependencies (infra/services):
  - <database / queue / 3rd-party API>
Configuration:
  - <env var / config key> — <purpose>
Test coverage:
  - <test file> — <what it covers>

---

Phase 2: Explore the Current (Target) Project

Use Serena for all current-project exploration. Do NOT use Bash ls/cat/find/grep on the current project's code files.

2a. Recall Context

• mcp__serena__list_memories → scan for memories related to the feature area and any architectural decisions
• mcp__serena__read_memory on relevant memories
• Read wiki/work/decisions/ index (if present) for accepted architectural decisions that constrain the port

2b. Architecture & Tech Stack

• mcp__serena__list_dir(".", recursive=false) — top-level shape
• mcp__serena__list_dir("<src-root>", recursive=false) — source structure
• Read package.json / equivalent for declared dependencies and scripts
• mcp__serena__get_symbols_overview on main source directories — identify modules, services, key interfaces

2c. Find Analogous Functionality

Search for anything in the current project that overlaps with or is adjacent to the feature being ported:

• mcp__serena__find_symbol with name patterns from the source feature inventory
• mcp__serena__search_for_pattern for key terms, library names, or data model names from the source
• Check wiki/work/tasks/, wiki/work/requirements/, and wiki/work/roadmaps/ for any planned or in-progress work in this area

2d. Build a Target Project Profile

Produce an internal summary:

TARGET PROJECT PROFILE
Language / Runtime: <language + version>
Primary frameworks:
  - <framework>: <role>
Relevant existing modules:
  - <module/file> — <what it does>
Existing analogous functionality:
  - <description of overlap or adjacent features>
Conventions observed:
  - File organisation: <pattern>
  - Naming: <pattern>
  - Error handling: <pattern>
  - Testing: <pattern>
  - State management / data access: <pattern>
Existing dependencies relevant to the port:
  - <dep>: <why relevant>
Missing dependencies (source uses, target does not have):
  - <dep>: <what source uses it for>

---

Phase 3: Gap Analysis & Adaptation Strategy

With both inventories in hand, perform a structured comparison.

3a. Technology Mapping

For each component of the source feature, determine the target equivalent:

3b. Complexity Classification

Classify each source file or module by porting effort:

Effort levels: Direct (copy + light edits) · Adapt (rework API calls, preserve logic) · Rewrite (logic must be re-expressed from scratch in target conventions)

3c. Dependency Assessment

For each dependency the source feature uses that the target project lacks:

Prefer using existing target project dependencies before recommending new ones.

3d. Integration Points

Identify where the ported feature will need to connect with existing target code:

• Entry points: Where does the feature get wired in (router, DI container, main module, CLI command, etc.)?
• Data layer: Does the feature need new tables/collections/indexes, or can it reuse existing ones?
• Auth / security: Does the feature touch auth — and how does the target project handle auth today?
• Config / env: What new env vars or config keys are needed?
• Tests: What test infrastructure does the target project use (jest, pytest, go test, etc.) and how should tests be structured?

3e. Risks & Constraints

List risks that could complicate the port:

---

Phase 4: Produce the Port Plan

4a. Write the Port Plan File

Write a structured port plan to .docs/port/<kebab-feature-name>.md:

---
source-project: <source-path>
feature: <feature name>
assessed: <YYYY-MM-DD>
effort: <Low / Medium / High / Very High>
status: draft
---

# Port Plan: <Feature Name>

> **Source**: `<source-path>`
> **Feature**: <functionality description>
> **Overall effort**: <Low / Medium / High / Very High>
> **Summary**: <One sentence on what this port involves.>

## Source Feature Summary

<Condensed version of the Source Feature Inventory from Phase 1c>

## Target Project Context

<Condensed version of the Target Project Profile from Phase 2d>

## Technology Mapping

<Table from 3a>

## Complexity Breakdown

<Table from 3b>

## Dependency Changes

<Table from 3c>

## Integration Points

<Bullet list from 3d>

## Risks

<Table from 3e>

## Porting Steps

Ordered, dependency-respecting steps. Each step is concrete and actionable.

### Step 1: <title>

**Effort**: <Direct / Adapt / Rewrite>
**Files affected**:
- `<target file path>` — <create / modify / delete>

**What to do**:
<Specific instructions: what to copy, what to change, how to adapt the API, what conventions to follow>

**Definition of done**: <How to know this step is complete>

---

### Step 2: <title>

… (repeat for each step)

---

## Suggested Next Actions

- To create a task from this plan: `/task-add "Port <feature name> from <source-path>" --roadmap <if applicable>`
- To explore the source further: use Read on `<most important source file>`
- If a dependency decision is needed: <frame the specific decision>

4b. Optionally Create a Task File

If the user requested a task file (or if the plan is actionable enough to execute), suggest:

/task-add "Port <feature name> from <source-project-name>"

And note what context from the port plan should be included in the task.

---

Phase 5: Report to User

After writing the plan file:

1. State the path written: .docs/port/<kebab-feature-name>.md 2. Give a one-line overall effort verdict (e.g., "Medium effort — 4 files to adapt, 1 rewrite, no new dependencies required") 3. List the top 3 porting steps in brief 4. Call out the single biggest risk or blocker 5. Suggest the most useful next action (task creation, further exploration, or a specific decision to make)

---

CRITICAL Rules

1. Serena for the current project only — all current-project code exploration must use Serena tools; no Bash ls/cat/find/grep on current project code files 2. Read + Bash for the source project — Serena cannot reach outside the current project; use Read and Bash to explore <source-path> 3. No speculation — every claim in the port plan must be grounded in what you actually read from either codebase 4. Prefer existing dependencies — do not recommend adding a new dependency if the target project already has something that covers the need 5. Respect target conventions — the port plan must describe porting the feature in the style and patterns of the target project, not a direct copy of the source 6. Brave Search only if needed — use Brave Search (sequential, 1 req/sec) only to compare alternative dependency options; do not use it to look up library docs (use Context7 instead)