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 skills add decodingai-magazine/building-a-coding-agent-from-scratch-course --skill demo-4-review-swarmgit clone --depth 1 https://github.com/decodingai-magazine/building-a-coding-agent-from-scratch-courseWrote 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/skills/decodingai-magazine/building-a-coding-agent-from-scratch-course/demo-4-review-swarm)<a href="https://agentmods.dev/skills/decodingai-magazine/building-a-coding-agent-from-scratch-course/demo-4-review-swarm"><img src="https://agentmods.dev/badge/skills/decodingai-magazine/building-a-coding-agent-from-scratch-course/demo-4-review-swarm/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/skills/decodingai-magazine/building-a-coding-agent-from-scratch-course/demo-4-review-swarm"><img src="https://agentmods.dev/badge/skills/decodingai-magazine/building-a-coding-agent-from-scratch-course/demo-4-review-swarm.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00050 | $0.00751 |
| Opus 5 | $0.00025 | $0.00376 |
| Sonnet 5 | $0.00010 | $0.00150 |
| Haiku 4.5 | $0.00005 | $0.00075 |
Grade A, and why
demo-4-review-swarm 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 11d 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 — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Run a parallel code-review swarm over decode's own source: three read-only Explore subagents review three modules at once, and you fold their reports into a single severity-ranked verdict with a health scorecard.
This is a showcase of the agent tool's native parallel fan-out (ADR-0013): N agent(...) calls
in one response run concurrently, each an Explore subagent whose toolset is read-only by
construction (read / glob / grep / lsp — no write, edit, or bash). Nothing the
subagents do mutates the repo.
Fan out — three subagents in ONE response
Spawn all three Explore subagents in a single model turn (so they run in parallel, not one after the other), one per module:
- Subagent A →
src/decode/permissions/— the allow/ask/deny gate, rules, and permission modes. - Subagent B →
src/decode/sandbox/— the executor seam, docker/modal backends, workspace, and hand-back. - Subagent C →
src/decode/context/— compaction and the JSONL session log.
Give each subagent the same brief: read every file in its module and report, as a read-only reviewer, on correctness, clarity, error handling, and any risky edges. Ask each one to return:
- A short prose summary of what the module does and how its files fit together.
- A text-based diagram of the module's structure — a Mermaid
flowchartshowing the files and how they call each other. - A flat list of findings, each tagged with a severity: Critical, Major, or Minor —
each naming the
file:lineit refers to. - A health score from 1 to 10 with a one-sentence justification.
Merge — one severity-ranked verdict
When all three reports come back, fold them into ONE verdict — do NOT paste the three reports verbatim:
- A health scorecard up top: one row per module with its score /10 and the one-line justification.
- A single findings table across all three modules, ranked by severity: every Critical
first, then every Major, then every Minor — each row naming the module and the
file:lineit refers to. - The three modules' Mermaid diagrams, one per module, each under its own heading.
- A two-or-three-line closing judgement: is the reviewed code healthy, and what is the single most important thing to fix first?
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.
- 11d ago First seen · 59 lines · 50 tokens per session scan A 44c1d832b0d3
demo-4-review-swarm is a skill published in the GitHub repository decodingai-magazine/building-a-coding-agent-from-scratch-course (383 stars, last pushed yesterday), licensed Apache-2.0. It adds 50 tokens to every session and 751 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.
Other skills, from other repositories
review
Review a change for correctness, security, and missing tests.
review-pr
Deeply review a GitHub pull request.
migration-review
Review database migration files when a change adds or modifies paths under migrations/. Use it before merge to collect forward, rollback, locking, and data-safety evidence.
michel-packmind-engineer-review
Review an implemented GitHub issue the way a senior Packmind engineer would — the human-judgment checks that ESLint, the TypeScript compiler, and e2e tests cannot catch (authorization scoping, hexagonal-architecture conformance, analytics-event wiring, UX copy, UI reactivity, CLI behavior, multi-tenancy safety, and…
idea-team
A structured group discussion for exploring a product idea through three roles: researcher, critic, and analogy finder. The roles look for evidence, weaknesses, and useful comparisons.
codex-agent
Use when you want a second-opinion review via Codex CLI, cross-verification after another agent implements changes, debugging help, or alternative implementation proposals. Requires Codex CLI to be installed and authenticated.