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 tolgakisaogullari/SumelaOS --skill requesting-code-reviewgit clone --depth 1 https://github.com/tolgakisaogullari/SumelaOSWrote 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/tolgakisaogullari/sumelaos/requesting-code-review)<a href="https://agentmods.dev/skills/tolgakisaogullari/sumelaos/requesting-code-review"><img src="https://agentmods.dev/badge/skills/tolgakisaogullari/sumelaos/requesting-code-review/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/tolgakisaogullari/sumelaos/requesting-code-review"><img src="https://agentmods.dev/badge/skills/tolgakisaogullari/sumelaos/requesting-code-review.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 | $0.00030 | $0.03177 |
| Opus 5 | $0.00015 | $0.01588 |
| Sonnet 5 | $0.00006 | $0.00635 |
| Haiku 4.5 | $0.00003 | $0.00318 |
Grade A, and why
requesting-code-review 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 9d 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 — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
<usage_criteria>
- MANDATORY: Before making ANY commit (must review staged/uncommitted changes first), after completing a task/feature, and before merging to main.
- SECURITY GATE: This skill acts as the final firewall. Code MUST be evaluated against security guidelines before it enters history.
- OPTIONAL (still valuable): when stuck on a problem (fresh perspective), before a major refactor (baseline check), or after fixing a complex bug.
- FORBIDDEN: Never skip reviews because a change seems "simple" or "small". </usage_criteria>
<dispatch_workflow> This skill dispatches a parallel review panel: a small set of lane reviewers run concurrently in isolated subagent contexts, each owning a distinct set of dimensions, then you synthesize their reports into one. Parallel + dimension-focused beats a single generalist reviewer — each lane goes deeper, and the lanes cover failure modes a single pass tends to miss. The panel is task-scoped (Step 3): a mandatory Correctness + Security floor ALWAYS runs, and you compose the remaining lanes to fit the change. Execute these steps strictly:
-
GATHER CONTEXT (CRITICAL PRE-COMMIT GATE):
- For staged changes: Get the diff of the currently staged files using
git diff --staged. - If the staged diff is empty but the task is complete, check
git diffbefore review so uncommitted work is not accidentally skipped. - For intentionally unstaged reviews, state that explicitly in
{HEAD_SHA}(for example, "Unstaged Working Tree") and pass the exactgit diffoutput as{CODE_DIFF}. - For already committed changes (if specifically requested): Get the exact git SHAs (e.g.,
BASE_SHA=$(git rev-parse HEAD~1)andHEAD_SHA=$(git rev-parse HEAD)).
- For staged changes: Get the diff of the currently staged files using
-
CHANGE SIZE CHECK (BEFORE DISPATCH): Count lines changed in the diff. If the diff exceeds ~300 lines of meaningful logic changes, STOP and ask the author to split the change. Reviewers cannot give quality feedback on 1000-line diffs; large reviews create rubber-stamping risk. Exception: Complete file deletions and automated refactoring where intent is obvious.
-
SCOPE THE PANEL (choose the lanes to fit THIS change):
- MANDATORY FLOOR — never skip: Lane 1 Correctness & Security ALWAYS runs (
requesting-code-review/reviewer-correctness-security.md). Security is non-negotiable regardless of how "trivial" or "docs-only" the change looks — a tiny change can still commit a secret, weaken auth, or break a contract. Do NOT drop this lane for any task. - COMPOSE the remaining 1–3 lanes to fit the change, drawing from the dimension catalog below. Total panel = 2–4 lanes (default 3). For a GENERAL code change, default to Lane 1 PLUS the two standard lanes (3 total:
reviewer-design-contracts.md+reviewer-integration-ops.md). For a SPECIALIZED change, replace or add a task-specific lane (author its prompt — see Step 5). - Dimension catalog — the starred (*) dimensions are ALWAYS Lane 1's, non-negotiable; the "relevance" filter applies only to the NON-starred ones: assign every non-starred dimension relevant to the diff to exactly one lane (don't drop a relevant one just because it doesn't fit a standard lane):
correctness*, security*, auth/credential-token-lifecycle*, conventions/readability, architecture, API/contract stability, backward-compat, integration/impact (graphify), performance, data/persistence & migrations, observability/rollback, testing (adaptive to TDD mode), concurrency/async/state, error-handling/resilience, i18n & accessibility (UI), prompt/LLM-cost (AI features), infra/IaC blast-radius. (* = the floor — always covered by Lane 1.)
- Note: Lane 1 already inspects the correctness facet of
concurrency/async/stateanderror-handling/resilience(race conditions, error paths). Spinning up a dedicated lane for these on a relevant change is fine — it DEEPENS that facet; the overlap with the floor is intentional, not an "exactly one lane" violation.
- Note: Lane 1 already inspects the correctness facet of
- Example task-specific lanes: schema/DB change → a "Data & Migration Safety" lane; heavy threading/async → a "Concurrency & State" lane; an AI/LLM feature → a "Prompt & Token-Cost" lane; pure UI → a "UX, i18n & Accessibility" lane; infra/Terraform → an "Infra & Blast-Radius" lane.
- Keep it 2–4 lanes: more fragments attention, fewer loses the cross-check. State the chosen lanes in one line so the user sees the composition (e.g., "Panel: Correctness & Security · Data & Migration Safety · Integration & Ops").
- MANDATORY FLOOR — never skip: Lane 1 Correctness & Security ALWAYS runs (
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 9d ago First seen · 97 lines · 30 tokens per session scan A d7481b9ac4bd
requesting-code-review is a skill published in the GitHub repository tolgakisaogullari/SumelaOS (4 stars, last pushed 15d ago), licensed MIT. It adds 30 tokens to every session and 3,177 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 skills, from other repositories
autoreview
Pre-commit/ship code review: Codex default; optional Claude or Pi.
rework-rate
Measure and interpret PR rework rate — the emerging 5th DORA metric.
omh-code-review
This is a Hermes-native code-review workflow skill.
revdiff-plan
Review the last Codex assistant message (plan, analysis, or proposal) with inline annotations in a TUI overlay. Extracts the most recent response from Codex rollout files and opens it in revdiff for review and annotation. Activates on "revdiff-plan", "review plan with revdiff", "annotate plan", "review last response"…
code-reviewer
Code review specialist focused on patterns, bugs, security, and performance.
agent-teams-simplify-and-harden
Implementation + audit loop using parallel agent teams with structured simplify, harden, and document passes. Spawns implementation agents to do the work, then audit agents to find complexity, security gaps, and spec deviations, then loops until code compiles cleanly, all tests pass, and auditors find zero issues or…