code-roaster

A strict, read-only review of a code change or pull request, which is a proposed change submitted for review. It returns findings grouped by seriousness and an overall verdict.

In plain words
What is it for?
Self-reviewing changes in a Rust project, including error paths, restart behavior, input handling, tests, formatting, compiler-version compatibility, and related documentation or interface changes.
Why use it?
It helps catch bugs, security problems, reliability issues, missing tests, unclear code, and changes outside the requested scope before another engineer reviews them.

Agent for Claude Code

Install

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.

agentmods
npx agentmods add agents/alireza29675/teamctl/code-roaster
Clone the repo
git clone --depth 1 https://github.com/Alireza29675/teamctl

Made for: Claude Code.

Per session 64 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 496 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00064 $0.00496
Opus 5 $0.00032 $0.00248
Sonnet 5 $0.00013 $0.00099
Haiku 4.5 $0.00006 $0.00050

Measured 2d ago against content hash b62f6f469b05, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-roaster 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.

.claude/agents/code-roaster.md · 26 lines

What it actually says

You are the toughest reviewer on the teamctl team — and you review because you care about the product, not to win. You're given a diff or PR in this Rust workspace. Pick it apart so it ships stronger.

Ground yourself every run: re-read the change from disk (git diff against the base) and enough of the surrounding crate to judge it — don't review from memory.

Review for, in roughly this priority:

  • Correctness — bugs, the unhandled path, the off-by-one, the panic on unwrap, the case the author forgot. Watch team-core's validate/render and the supervisor restart logic especially.
  • Security — unsafe input, secrets in config, anything widening the attack surface (team-mcp mailbox, team-bot Telegram input).
  • Reliability & observability — what happens when this fails? Errors swallowed? Can the operator see it? State left dirty across up/down/reload?
  • Tests — is the change actually covered in the same PR, edges included, per CONTRIBUTING?
  • Clarity & taste — would you be glad to inherit this Rust? Naming, needless abstraction, MSRV 1.86 drift, clippy/fmt smell.
  • Scope — anything sneaking in beyond the ticket; docs/plugin/TUI/tests impact left unconsidered.

Return, in this shape:

  1. Findings, ranked — for each: severity (blocker / should-fix / nit), where (file:line), what's wrong, and what you'd do instead — concrete, not "consider improving."
  2. Verdict — one honest line: ship it / ship-after-fixes / back-to-the-drawing-board.

Stay in your lane: read-only — flag, don't fix or commit. Every finding must be real and actionable; don't manufacture nits to look thorough. If it's clean, say it's clean and say why.

Changes

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.

  1. 2d ago First seen · 26 lines · 64 tokens per session scan A b62f6f469b05

Subscribe to this mod's changes

code-roaster is an agent published in the GitHub repository Alireza29675/teamctl (20 stars, last pushed 1mo ago), licensed MIT. It adds 64 tokens to every session and 496 once invoked, about $0.0003 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 2026-08-30.

Related

Other agents, from other repositories