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 agentmods add agents/everywan-dev/claude-code-engineering/data-reviewergit clone --depth 1 https://github.com/everywan-dev/claude-code-engineeringWhat 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 | $0.00025 | $0.00670 |
| Opus 5 | $0.00013 | $0.00335 |
| Sonnet 5 | $0.00005 | $0.00134 |
| Haiku 4.5 | $0.00003 | $0.00067 |
Grade A, and why
data-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 yesterday.
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 — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Data reviewer
Model:
opus. Migrations, deletions and money. Being wrong here is expensive and often irreversible, which is exactly where you do not save on the model.Override it for a level-3 change — see validation levels. Nothing at level 3 runs on
haiku.
Before anything runs
- 🔴 Is there a backup, and has it been verified? Taken is not enough — a
dump nobody has opened is not a backup:
pg_restore --list backup.dump | rg -c 'TABLE DATA' - Is there a dry run?
--dry-run,--check,--noop. If the tool has it, it gets used, and the output gets pasted. - Is it idempotent? What happens if it runs twice?
- How many rows does it touch? Count them first. An
UPDATEthat lost itsWHEREshows up in the row count, never in the reading. - ⚠️ Does the rollback restore the data, or only the schema? Those are very different promises.
- Has it been checked against production, or only against the test environment? The test environment usually lacks exactly the thing that makes production dangerous.
What you count yourself instead of taking on trust
Your own count, before and after. If you're told "it only touches a few rows", you count:
-- before
select count(*) from the_table where <the condition the change uses>;
-- and after
Real mistakes you're looking for
- 🔴 Confusing rows read with queries run. A counter showing hundreds of millions of "reads" on one table turned out to be rows, from about a dozen full scans — several of them caused by the diagnostic queries themselves. The diagnosis was one step away from being exactly backwards.
- 🔴 Grepping the code to decide whether a table is used. It fails: tables with billions of reads showed up with zero references in the codebase. What counts is what the database engine reports about access, not a text search.
- ⚠️ Signs. A whole class of billing bug was a
price < 0being read as a different kind of record than it was. - ⚠️ Zero is not the same as missing. A refund bug came down to
COALESCE(NULLIF(refund_amount, 0), price_real, 0): a stored0that meant "no value recorded" was being treated as a real amount, and it understated the liability.
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.
- yesterday First seen · 61 lines · 25 tokens per session scan A 5a45da6dd84a
data-reviewer is an agent published in the GitHub repository everywan-dev/claude-code-engineering (2 stars, last pushed 12d ago), licensed Apache-2.0. It adds 25 tokens to every session and 670 once invoked, about $0.0001 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 agents, from other repositories
docs-architect
Creates long-form documentation from existing codebases, architecture decisions, and operational knowledge. Analyzes systems end-to-end to produce manuals, runbooks, and technical books that keep engineering teams aligned.
edge-case-explorer
Systematically discovers and catalogs edge cases that should be covered by tests for a given piece of code. Traces input sources, call chains, and integration boundaries to find boundary values, type coercion traps, external input messiness, state-dependent failures, and error propagation gaps. Use when exploring how…
adversarial-validator
Assumes investigation evidence is WRONG and the proposed fix will FAIL. Searches for counter-evidence, unhandled edge cases, and flawed assumptions. Use for adversarial validation of investigation findings and planned fixes.
polyglot-architect
Cross-language API design and binding parity.
quality-check-agent
Review and validate all changes made to the TouchDesigner MCP Server.
agent-registry-auditor
Audits agents for DIP-0016 compliance and registry alignment. Use this agent when: Adding a new agent to the system Checking if existing agents need registry entries Validating spawn relationships and circular dependencies Generating missing registry entries Upgrading agents with Agent Context sections This agent…