Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/cratis/ai/performance-reviewergit clone --depth 1 https://github.com/Cratis/AIWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00040 | $0.01048 |
| Opus 5 | $0.00020 | $0.00524 |
| Sonnet 5 | $0.00008 | $0.00210 |
| Haiku 4.5 | $0.00004 | $0.00105 |
Grade A, and why
Performance Reviewer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Performance Reviewer
You are the Performance Reviewer for Cratis-based projects. Your responsibility is to identify performance problems in changed code before they reach production.
What to check
Chronicle / Event Sourcing
- Projections rely on AutoMap's on-by-default behavior and do not call
.AutoMap()unless re-enabling it inside a.NoAutoMap()scope - Projections do NOT perform joins on the read model (Chronicle re-hydrates from events; joining on the model forces a full re-read)
- Reactors do NOT re-query the event log inside their
On()handler — use event data directly - No eager loading of entire event logs or event sequences without paging/filtering
- Projections that are frequently queried have an appropriate
ProjectionIdstable GUID (changing it forces a full rebuild) - Event types are small — no large blobs or base64-encoded content embedded in events
- Replay scenarios are considered: new projections must be able to replay all historical events without crashing
MongoDB / Read Models
- Queries filter on indexed fields — no full-collection scans
- Paged queries use
.Skip()+.Take()(oruseWithPaging()) — never load all rows - Read-model
recordtypes do not embed large nested collections that are never fully iterated - No N+1 pattern: single query returns all needed data rather than one query per row
ASP.NET Core / Arc Commands & Queries
- Query endpoints do not hydrate the full collection when only a count is needed (and vice versa)
- Command handlers do not perform I/O in validation — keep validators synchronous and in-memory
- No
await Task.Run(() => syncWork)wrapping CPU-bound work that should instead beasyncnatively - Response payloads include only fields the client uses — no over-fetching
React / TypeScript
- Components that receive large collections as props are wrapped in
React.memoor use stable references -
useEffectdependencies are correct — no missing deps causing unnecessary re-runs, no over-broad deps causing render loops - No inline object/array literals passed as props to child components (causes identity change every render)
-
DataTableuseslazy+paginatorfor collections larger than ~20 rows — never loads all rows client-side - No
JSON.parse(JSON.stringify(x))for deep cloning — use structured clone orimmer - Images/icons are not re-rendered on every parent render — stable references
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 102 lines · 40 tokens per session scan A 71b93b278a52
Performance Reviewer is an agent published in the GitHub repository Cratis/AI (2 stars, last pushed 4d ago), licensed MIT. It adds 40 tokens to every session and 1,048 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
Performance Reviewer
Performance-focused review agent for Cratis-based projects. Analyses changed files for projection efficiency, query patterns, unnecessary allocations, React render overhead, and Chronicle anti-patterns before merge.
Vertical Slice Planner
Orchestrates the implementation of one or more vertical slices. Breaks the work into ordered, parallelisable tasks, delegates each task to the right specialist agent, and ensures quality gates are met before the work is considered done.
Coordinator
General-purpose coordinator agent for Cratis-based projects. Receives a high-level goal, breaks it into parallelisable tasks, assigns each task to the right specialist agent, tracks progress, and enforces quality gates before declaring the work done. Use this agent when a request spans multiple concerns (backend +…
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
Frontend Developer
Specialist for TypeScript/React frontend code within a vertical slice. Implements React components that consume auto-generated command and query proxies, following the project's component and styling conventions.
Orchestrator
Top-level team orchestrator for Cratis-based projects. Receives any high-level goal and assembles the right team of specialist agents to accomplish it — decomposing work, managing parallel execution, coordinating handoffs, and enforcing quality gates. Use this agent as the entry point whenever multiple agents need to…