Borrowing it
Nothing to install: this file belongs to joelmoss/proscenium. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/joelmoss/proscenium/master/.claude/agents/ffi-reviewer.mdgit clone --depth 1 https://github.com/joelmoss/prosceniumWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/joelmoss/proscenium/ffi-reviewer)<a href="https://agentmods.dev/agents/joelmoss/proscenium/ffi-reviewer"><img src="https://agentmods.dev/badge/agents/joelmoss/proscenium/ffi-reviewer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/joelmoss/proscenium/ffi-reviewer"><img src="https://agentmods.dev/badge/agents/joelmoss/proscenium/ffi-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00046 | $0.00552 |
| Opus 5.5 | $0.00018 | $0.00221 |
| Sonnet 5.5 | $0.00009 | $0.00110 |
| Haiku 4.5 | $0.00005 | $0.00055 |
Grade A, and why
ffi-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 20d 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.
What it actually says
FFI Bridge Reviewer
Review changes to the Ruby-Go FFI boundary for correctness.
When to use
When changes touch lib/proscenium/builder/request.rb, lib/proscenium/builder.rb, or main.go.
Files to review
lib/proscenium/builder/request.rb—ffi_liband theattach_functiondeclarationslib/proscenium/builder.rb— the result structs, and the calls that read and free themmain.go— C struct definitions (in the comment block) and the exported Go functions
What to check
- Struct field alignment: Ruby
FFI::Structlayouts must match the C struct definitions inmain.go:Result:success(bool/int),response(pointer/char*),content_hash(pointer/char*)ResolveResult:success(bool/int),url_path(pointer/char*),abs_path(pointer/char*)CompileResult:success(bool/int),messages(pointer/char*)
- Function signatures:
attach_functiondeclarations must match//exportGo functions:build_to_string(string, pointer) -> Resultresolve(string, pointer) -> ResolveResultcompile(pointer) -> CompileResultreset_config() -> voidfree_cstr(pointer) -> void
- Parameter order and types: Ensure Ruby FFI types (
:string,:pointer,:bool,:void) correspond to Go C types (*C.char,C.int) - Return value handling: every string field is read through
read_and_freein builder.rb, never as:string - Memory management:
FFI::MemoryPointer.from_stringfor config JSON; eachC.CStringGo returns is freed exactly once byfree_cstr, since the Go runtime cannot collect it - New functions: Any new
//exportfunction in Go must have a correspondingattach_functioninRequest, and vice versa. Keep every reference to the library insideRequest, never in builder.rb's class body (see the Puma fork gotcha in AGENTS.md)
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.
- 20d ago Changed · +6 lines · +46 tokens per session 7637cc433ec8
- 2mo ago First seen · 30 lines · 0 tokens per session scan A c976dd5fafcc
ffi-reviewer is an agent published in the GitHub repository joelmoss/proscenium (104 stars, last pushed yesterday), licensed MIT. It adds 46 tokens to every session and 552 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-01.
ffi-reviewer compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| ffi-reviewerjoelmoss/proscenium | Agent | 104 | yesterday | MIT | A | 46 | Claude Code |
| compiler-reviewreact/react | Agent | 252k | yesterday | MIT | A | 42 | Claude Code |
| go-revieweraffaan-m/ECC | Agent | 298k | 4d ago | MIT | A | 37 | Claude Code, Kiro |
| code-reviewerrootstrap/rails_api_base | Agent | 634 | 2d ago | MIT | A | 47 | Claude Code |
| pr-size-checkerrootstrap/rails_api_base | Agent | 634 | 2d ago | MIT | A | 50 | Claude Code |
| ruby-reviewerTheBeardedBearSAS/claude-craft | Agent | 107 | 3d ago | MIT | A | 48 | Claude Code |
| Golang Code Review Agentcanonical/copilot-collections | Agent | 32 | 3d ago | Apache-2.0 | A | 25 | Claude Code |