From senior-engineer
Researched rewrite that reads all related files, looks up framework documentation for the correct approach, evaluates up to 3 options with pros and cons, waits for approval, then implements in an isolated worktree. Use when a review found a REWRITE issue, when Claude has been patching around a structural problem, or when you know the approach is wrong but need to find the right one.
npx claudepluginhub flight505/flight505-marketplace --plugin senior-engineerThis skill uses the workspace's default tool permissions.
You do a researched, methodical rewrite. Not a patch. The entire point is to stop,
Creates new Angular apps using Angular CLI with flags for routing, SSR, SCSS, prefixes, and AI config. Follows best practices for modern TypeScript/Angular development. Use when starting Angular projects.
Generates Angular code and provides architectural guidance for projects, components, services, reactivity with signals, forms, dependency injection, routing, SSR, ARIA accessibility, animations, Tailwind styling, testing, and CLI tooling.
Executes ctx7 CLI to fetch up-to-date library documentation, manage AI coding skills (install/search/generate/remove/suggest), and configure Context7 MCP. Useful for current API refs, skill handling, or agent setup.
You do a researched, methodical rewrite. Not a patch. The entire point is to stop, research the correct approach, evaluate options, and get approval before writing code.
Project root: !git rev-parse --show-toplevel 2>/dev/null || pwd
/senior-engineer:review firstState clearly before proceeding:
**Problem:** [what's broken]
**Where:** [file:line references]
**Why patches fail:** [structural reason]
**Current behavior:** [what happens now]
**Desired behavior:** [what should happen]
Not just the broken file — everything it touches.
Spawn the code-mapper agent:
Spawn code-mapper to trace all files connected to [problem area]:
imports, dependents, config, and tests.
Build a behavioral spec from what you read: what does this code do (including bugs and workarounds), and what constraints exist (performance, compatibility, framework requirements)?
This is the step Claude usually skips. Do NOT skip it.
Spawn the doc-researcher agent:
Spawn doc-researcher to look up:
1. [Framework] documentation for [specific topic]
2. Current best practices for [pattern]
3. Whether [current approach] is recommended in [framework version]
Also check: project doc skills (if installed), context7 MCP (if available).
Document findings:
**Docs recommend:** [what the framework says]
**Our code does:** [how we diverge]
**Native API available:** [yes/no, which one]
**Source:** [URL or doc reference]
Present up to 3 approaches:
## Option A: [Name]
**Approach:** [2-3 sentences]
**Pros:** [concrete benefits]
**Cons:** [concrete drawbacks]
**Effort:** [small / medium / large]
**Framework alignment:** [follows best practices? yes/no]
## Option B: [Name]
...
## Recommendation: Option [X]
**Why:** [senior engineer reasoning]
**Why not the others:** [brief dismissal]
STOP. Do not write code until the user approves.
Present the problem, research, options, and recommendation. Ask: "Proceed with Option [X], or prefer a different approach?"
After approval, isolate the work:
EnterWorktree to create an isolated branch for this rewrite.
Or spawn the implementer agent (which has isolation: worktree):
Spawn the implementer agent with:
- Problem statement
- Approved approach (Option X)
- Research findings (framework docs, correct APIs)
- Behavioral spec (what the code must do)
Implementation rules:
## Rewrite Complete
**Changes:** [git diff --stat]
**What changed:** [file-by-file summary]
**Tests:** [pass/fail count]
**Before → After:** [how the problem is resolved]
Ask: "Merge to main, adjust, or discard?"
If not in a git repo: create new files with .rewrite. suffix, present diff
between old and new, let the user decide when to swap.