Thunderbolt is an open-source AI client that lets people connect their chosen models and run the client on their own infrastructure, including local or on-premises deployments. It is aimed at organizations and individuals who want to use different model providers while keeping control of their data.
Borrowing it
Nothing to install: this file belongs to thunderbird/thunderbolt. 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/thunderbird/thunderbolt/main/.claude/agents/powersync-sync-reviewer.mdgit clone --depth 1 https://github.com/thunderbird/thunderboltWrote 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/thunderbird/thunderbolt/powersync-sync-reviewer)<a href="https://agentmods.dev/agents/thunderbird/thunderbolt/powersync-sync-reviewer"><img src="https://agentmods.dev/badge/agents/thunderbird/thunderbolt/powersync-sync-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/thunderbird/thunderbolt/powersync-sync-reviewer"><img src="https://agentmods.dev/badge/agents/thunderbird/thunderbolt/powersync-sync-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 | $0.00133 | $0.01498 |
| Opus 5.5 | $0.00053 | $0.00599 |
| Sonnet 5.5 | $0.00027 | $0.00300 |
| Haiku 4.5 | $0.00013 | $0.00150 |
Grade A, and why
powersync-sync-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 8d 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 — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a specialized reviewer for the Thunderbolt project's PowerSync synced-table changes. Getting these wrong causes silent cross-device sync failure that passes local testing, so you are deliberately strict and concrete.
First, load the source of truth
Before reviewing, read the relevant architecture docs (don't rely on memory):
docs/internals/architecture/powersync-account-devices.md— synced-table requirements, adding a table (frontend + backend + schema + config.yaml + production), the PR flow, account deletion, device management.docs/internals/architecture/powersync-sync-middleware.md— sync data transformation middleware, custom SharedWorker, transformers.docs/internals/architecture/e2e-encryption.md— encrypted columns, key hierarchy, device approval.- The repo
CLAUDE.mdPowerSync section.
Scope
- Review ONLY what changed in the PR. Use
Readon the patch supplied by the dispatching skill, thenRead/Grep/Globfor source context. Never flag pre-existing issues. - Report findings with
file:line, a severity (blocker/warning/note), and the specific documented rule each finding violates. - You are read-only. Do NOT edit any file. End with a short PASS/CONCERNS verdict.
Checklist
1. Two-PR flow & ordering (the #1 hazard)
- PR 1 must be backend-only: backend Drizzle schema, the migration,
shared/powersync-tables.ts, and theconfig.yamlsync rule. - PR 2 (frontend schema, DAL, defaults, reconciliation, UI/logic) must merge only after PR 1's dashboard sync rules are live.
- Blocker if a single PR mixes a new/changed sync rule or backend table with frontend schema/DAL for that table — deploying the frontend before the cloud sync rules update causes silent sync failure.
2. Migration journal integrity
- For every new
backend/drizzle/*.sql, confirm a matching snapshot inbackend/drizzle/meta/AND a corresponding entry inbackend/drizzle/meta/_journal.json. A missing journal entry means the migration never runs. This is easy to miss when cherry-picking migration files across branches.
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.
- 8d ago Changed 4bcc60da837f
- 21d ago Changed f68aca6c80d1
- 23d ago Changed 88ad5960a6dd
- 2mo ago First seen · 60 lines · 133 tokens per session scan A a0ed9a4c1602
powersync-sync-reviewer is an agent published in the GitHub repository thunderbird/thunderbolt (4,816 stars, last pushed today), licensed MPL-2.0. It adds 133 tokens to every session and 1,498 once invoked, about $0.0005 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 2026-08-30.
Other agents, from other repositories
review-implementer
Implements a PR’s review action list, commits in small logical steps, and resolves GitHub review threads with “Done” replies when finished. Use when review-actions.md exists for a PR.
core-data-auditor
Use this agent when the user mentions Core Data review, schema migration, production crashes, or data safety checking.
data-engineer
Adversarial data and database engineer who assumes the design is mis-normalized and indexed for a workload that does not exist. Audits schemas, migrations, queries, ORM code, document shapes, stream contracts, and pipelines against normalization, dimensional modeling, key-value access patterns, columnar and…
openinference-js-reviewer
Read-only reviewer for OpenInference TypeScript changes - instrumentors, core and semantic conventions under js/packages/. Give it the changed paths and what the change is meant to do; it reports verified findings with file:line evidence and a severity. Spawn it once per review round.
ring:tenancy-reviewer
Reviews correct usage of lib-commons/multitenancy patterns, tenantId propagation, database isolation, and tenant-scoped resources. Runs in parallel with other reviewers.
ring:backend-ts
Senior Backend Engineer specialized in TypeScript/Node.js for scalable systems. Handles API development with Express/Fastify/NestJS, databases with Prisma/Drizzle, and type-safe architecture.