メニュー
Audit the MCP server experience from an agent's perspective. Use when checking search quality, response richness, or first-impression for new MCP installs.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Translate git history and orchestrator logs into authentic social content. Use when turning recent technical work into tweets, Dev.to posts, or changelog entries without manual summarisation.
Red team the staging environment with injection attacks, rate limit probes, and auth bypass attempts. Use before major releases or when adding new user-input surfaces.
Deploy Your Project to production safely. Use before any deployment, after code changes, or when asked to push to prod.
Facilitate a structured CONVERSATIONAL meeting between department agents via claude-peers — named phases (Diverge, Clarify, Challenge, Build, Stress Test, Converge, Decide) with state machine navigation. Agents push back on each other and build on each other's ideas. Minimum path: Diverge → Challenge → Build → Stress Test. Early exit only after Stress Test. NOT for simple task dispatch (use orchestra-management). NOT for solo research (use rd-architect).
Manage the 6-department orchestra via claude-peers. Use when dispatching tasks, coordinating departments, reviewing results, or shaping department behavior.
Write emails, social posts, and newsletter submissions for Your Project. Use when creating maker outreach, blog content, social media posts, or press pitches.
| name | agent-experience-audit |
| description | Audit the MCP server experience from an agent's perspective. Use when checking search quality, response richness, or first-impression for new MCP installs. |
| metadata | {"version":"1.0.0","author":"Master Agent","category":"product-quality","updated":"2026-03-30T00:00:00.000Z"} |
You audit what AI agents see when they use Your Project's MCP server. Your goal is to ensure every interaction is obviously better than a raw web search.
Test the top 5 search queries (auth, analytics, payments, email, database). For each: is the top result relevant? Does it have install_command? Is the tagline useful?
Pick specific tools (e.g. the 13 we emailed) and check their full detail response. Are descriptions accurate? Are migration signals present? Are compatible tools shown?
Simulate a developer who just installed from Smithery/Glama. What's their first query? What do they see? Would they keep using us?
| Check | Pass | Fail |
|---|---|---|
| Top result is relevant to query? | Auth → auth tool | Auth → Airflow |
| Top result has install_command? | npm install X | Empty |
| Top result has useful tagline? | Describes what it does | Truncated or generic |
| Category is correct? | Analytics → Analytics | Analytics → AI & Automation |
| Migration signal present? | "14 repos migrated from X" | Empty (acceptable if no data) |
| Check | Pass | Fail |
|---|---|---|
| Description > tagline? | Full paragraph | Same as tagline |
| Install command correct? | Matches actual package | Wrong package entirely |
| Migration data present? | Gaining/losing signals | Empty (check sdk_packages mapping) |
| Verified combos present? | "Works with X in N repos" | Empty |
| Trust tier shown? | "New/Tested/Verified" | Missing |
# Search test
~/.fly/bin/fly ssh console -a your-project -C 'python3 -c "
import urllib.request, json
r = urllib.request.urlopen(\"http://localhost:8080/api/tools/search?q={query}&limit=5\")
data = json.loads(r.read())
for t in data.get(\"tools\", []):
print(t.get(\"slug\") + \": \" + str(t.get(\"migration_signal\", \"no signal\")))
"'