review-pro-synthesize

review-pro-synthesize is a skill for Claude Code from tufantunc/review-pro. It costs 45 tokens per session (2,591 once invoked), scanned A, original, MIT.

A final review stage that combines findings from several specialist code reviewers into one report. It removes duplicate findings, resolves disagreements, checks evidence, and assigns calibrated severity.

In plain words
What is it for?
Use it after specialist reviews to merge their findings, identify issues reported by multiple reviewers, verify evidence outside changed files, and produce prioritized remediations.
Why use it?
It prevents multiple reviewers from producing a confusing pile of repeated or conflicting comments. It leaves the team with one prioritized verdict and clearer fixes.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the review-pro plugin — 14 skills, 15 agents shipped together

Good fit Use it after specialist reviews to merge their findings, identify issues reported by multiple reviewers, verify evidence outside changed files, and produce prioritized remediations.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/tufantunc/review-pro/review-pro-synthesize
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.

Any agent
npx skills add tufantunc/review-pro --skill review-pro-synthesize
Clone the repo
git clone --depth 1 https://github.com/tufantunc/review-pro

Made for: Claude Code.

Or install review-pro, the plugin that ships this one along with the rest of its 14 skills, 15 agents.

Wrote 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.

agentmods badge for review-pro-synthesize

README.md
[![agentmods](https://agentmods.dev/badge/skills/tufantunc/review-pro/review-pro-synthesize/github.svg)](https://agentmods.dev/skills/tufantunc/review-pro/review-pro-synthesize)
Your own site
<a href="https://agentmods.dev/skills/tufantunc/review-pro/review-pro-synthesize"><img src="https://agentmods.dev/badge/skills/tufantunc/review-pro/review-pro-synthesize/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.

agentmods 80×15 button for review-pro-synthesize

Your own site · 80×15
<a href="https://agentmods.dev/skills/tufantunc/review-pro/review-pro-synthesize"><img src="https://agentmods.dev/badge/skills/tufantunc/review-pro/review-pro-synthesize.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 45 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,591 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00045 $0.02591
Opus 5 $0.00023 $0.01295
Sonnet 5 $0.00009 $0.00518
Haiku 4.5 $0.00005 $0.00259

Measured 12d ago against content hash e2b57cd9b744, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

review-pro-synthesize 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 12d 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.

core/skills/review-pro-synthesize/SKILL.md · 141 lines

How it starts

The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Review-Pro Synthesis (Stage 3)

You are the orchestrator's final stage. You receive the structured findings from all dispatched reviewers, plus diff_class, changed_files, spec_source, external_premises, and premises_dropped from triage's dispatch plan, and produce ONE unified review.

Steps

  1. Collect all finding blocks from the dispatched reviewers.
  2. Dedup by (file, line±5, category-root, overlap_hints): the same issue flagged by multiple reviewers collapses into one.
  3. Weight: a finding flagged by ≥2 reviewers gets a conviction boost — annotate it "flagged by N reviewers".
  4. Resolve conflicts by ownership — the domain owner sets severity (see table).
  5. Calibrate severity: enforce the anti-overreporting bar. Downgrade anything not fully traced to evidence. Never upgrade beyond what a specialist justified.
  6. Out-of-diff evidence check (see below) — a review-level confidence signal, not a per-finding gate.
  7. Verdict + prioritized findings + remediations.

Out-of-diff evidence check

Count the code-axis findings only whose evidence_refs name at least one path not in triage's changed_files: a caller, an existing guard, a canonical helper, a schema, an upstream source. A finding with no evidence_refs does not count toward the total.

Spec-axis findings are excluded from this count and it is not a detail. A spec finding's evidence is the spec document or issue, which is outside the diff in the issue and pr-body cases and may well be inside it for a file source (a design doc committed alongside its implementation). Either way the exclusion holds: counting spec findings would satisfy this check on most reviews where the axis ran and quietly disable it.

If triage reported diff_class: substantive and that count is zero, append this caveat to the report, immediately under the verdict:

> No finding in this review cites evidence outside the diff. For a change of this
> size that usually means the review never left the diff — treat the verdict with
> reduced confidence.

Read the full file on GitHub · 141 lines

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. 12d ago First seen · 141 lines · 45 tokens per session scan A e2b57cd9b744

Subscribe to this mod's changes

review-pro-synthesize is a skill published in the GitHub repository tufantunc/review-pro (4 stars, last pushed 3d ago), licensed MIT. It adds 45 tokens to every session and 2,591 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.

Related

Other skills, from other repositories

juror-review

Inspect Juror Cloud PR findings and, only after an explicit confirmation, start or rerun a hosted Juror review.

Juror-AI/juror · 28 tokens

logic-health

Sweep a directory, module, or full codebase for logic correctness and produce a scored health dashboard with systemic patterns. Trigger when the user requests a health view — "audit the whole codebase", "health check", "health overview", "logic health overview", "audit src/", "audit auth and payments modules", "where…

hyhmrright/logic-lens · 180 tokens

logic-diff

Compare two code versions for semantic equivalence via semi-formal tracing of both versions side-by-side. Trigger when the user shares a refactor, rewrite, migration, or A/B implementation and wants to confirm behavior is unchanged — "did I break anything", "is this equivalent", "are these equivalent", "semantically…

hyhmrright/logic-lens · 192 tokens

river-review-code

A general code-review skill that checks readability, maintainability, and type safety, with routing for several specialized checks.

s977043/river-review · 51 tokens

deep-review

Multi-angle code review that adapts to what is being reviewed — a mid-flight worktree increment, a full PR before production, a merge-window integration sweep, or a security-only pass. Findings are scored, adversarially refuted, widened to the same defect elsewhere in the codebase, and reported to the native review…

Jmosier69/refute · 97 tokens

Standard Review Policy for Midstream

Applies standard AI review policy guidelines for midstream (implementation) phase reviews.

s977043/river-review · 23 tokens