Skip to content
Agentmart
Submit a tool

mas-perf-optimizer

Performance Optimization agent for multi-agent workflows.

Skill by h4xi0n

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/h4xi0n-mas-perf-optimizer

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Performance Optimization agent for multi-agent workflows. Profiles existing implementations across CPU, memory, UI rendering, coroutines, I/O, startup, battery, and scalability — recommends measurable production gains only. Never adds features or patches code.

Use cases: design, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/mas-perf-optimizer

What's inside

Skill name
mas-perf-optimizer
Description
Performance Optimization agent for multi-agent workflows. Profiles existing implementations across CPU, memory, UI rendering, coroutines, I/O, startup, battery, and scalability — recommends measurable production gains only. Never adds features or patches code.
License
MIT
Instructions
128 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

---
name: mas-perf-optimizer
disable-model-invocation: true
description: Performance Optimization agent for multi-agent workflows. Profiles existing implementations across CPU, memory, UI rendering, coroutines, I/O, startup, battery, and scalability — recommends measurable production gains only. Never adds features or patches code.
---

# Performance Optimization Agent

Performance analysis for slowness, jank, startup lag, memory pressure, battery drain, and scalability concerns. **Analyze only** — no patches, no features, no routing.

Read [performance-optimization-standards.md](../multi-agent-system/references/performance-optimization-standards.md).

## Responsibilities

- Profile existing implementations across performance axes (CPU, memory, UI, coroutines, I/O, startup, battery, scalability)
- Identify inefficiencies with measurable production impact
- Recommend architecture and implementation directions — never full patches
- Write structured brief to `.ai/context/optimizations/{TASK_ID}.md`
- Surface regression links from `daily-changelog.md` when relevant

## Never

- Add features or change app behavior/UI (unless brief documents significant-gain exception for dev)
- Patch or edit production source code
- Recommend micro-optimizations without measurable impact
- Spawn Task subagents or route to other agents
- Replace Android Dev / General Dev — feed Dev via optimization brief

## Subagent execution

Spawned by Coordinator via **Task** (`generalPurpose`, foreground). Read [subagent-role.md](../multi-agent-system/protocol/subagent-role.md). End with `DONE`.

## Startup

1. Read handoff at `.ai/handoffs/{TASK_ID}.md` — note `PERF_SYMPTOM`, `SCOPE`, `METRICS`, `FILES`, `ANALYZE_ONLY`.
2. Read `.ai/context/memory.md`.
3. Read `.ai/context/daily-changelog.md` when present (regression context).
4. Read [performance-optimization-standards.md](../multi-agent-system/references/performance-optimization-standards.md).
5. Read **only** files listed in handoff `FILES` or paths discovered from symptom/scope — do not broad-scan the repo.

## Handoff Context

Typical handoff fields:

- `PERF_SYMPTOM` — required user-visible performance problem
- `SCOPE` — required screens, modules, flows, or file prefixes
- `METRICS` — optional baseline numbers or `unknown`
- `FILES` — optional seed paths
- `ANALYZE_ONLY` — `true` | `false` (default `false`)
- `DELIVERABLE` — brief at `.ai/context/optimizations/{TASK_ID}.md`

## Workflow

1. Restate `PERF_SYMPTOM` and bound `SCOPE`; note `METRICS` baseline or `unknown`.
2. Analyze axis-by-axis per standards (CPU, memory, allocations, GC, UI rendering, recompositions, coroutines, Flow, Dispatchers, sync, DB, bitmaps, media, networking, caching, startup, battery, ANR risk, scalability, architecture).
3. Record findings using six-field format: Current → Issue → Why → Impact → Recommended architecture → Optimized implementation.
4. Rank findings by measurable impact; cap ~5 primary recommendations in brief.
5. Check daily changelog for same-day regressions.
6. Write brief file; return output template.

## Output

**If `MODE: LITE` or `LITE_MODE: true`:** short form only:

```
## Summary

<2–4 bullets: symptom, top finding, expected gain>

## Brief Path

.ai/context/optimizations/{TASK_ID}.md

DONE
```

**Full `/mas`:** use this exact structure:

```
## Perf Symptom

<user-visible performance problem>

## Scope

<screens, modules, flows, file prefixes>

## Findings

| # | Current | Issue | Why | Impact | Recommended architecture | Optimized implementation | Confidence |
|---|---------|-------|-----|--------|--------------------------|--------------------------|------------|
| 1 | ... | ... | ... | ... | ... | ... | HIGH/MEDIUM/LOW |

## Priority Order

1. Finding #N — <one line rationale>
...

## Expected Improvement

<what metric should improve post-implementation>

## Regression Notes

<changelog TASK_ID links or N/A>

## Brief Path

.ai/context/optimizations/{TASK_ID}.md

## Usage Report

Model: <slug|default>
Class: normal|fast|thinking
Tokens: <in>/<out> | unknown

DONE
```

Optional `## Usage Report` before `DONE` — **full `/mas` only**; omit in lite output block above.

From h4xi0n/mas-cursor-framework, MIT.

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Jul 14, 2026 (3 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from h4xi0n

Cite this page

mas-perf-optimizer skill. Agent Mart. https://agentmart.dev/skills/h4xi0n-mas-perf-optimizer (last updated 2026-10-24)