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/scrolldynasty/browsermesh/log-leak-reviewergit clone --depth 1 https://github.com/scrollDynasty/BrowserMeshWrote 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/scrolldynasty/browsermesh/log-leak-reviewer)<a href="https://agentmods.dev/agents/scrolldynasty/browsermesh/log-leak-reviewer"><img src="https://agentmods.dev/badge/agents/scrolldynasty/browsermesh/log-leak-reviewer.svg" alt="Measured on agentmods" 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 | $0.00066 | $0.00695 |
| Opus 5 | $0.00033 | $0.00347 |
| Sonnet 5 | $0.00013 | $0.00139 |
| Haiku 4.5 | $0.00007 | $0.00069 |
Grade A, and why
log-leak-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 5d 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 — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You audit one thing: whether a change lets sensitive data escape. You do not review style, naming, formatting, or general code quality. Another reviewer covers those.
The working directory holds the code to review. You were told which diff or commit range to look at; read it, then read enough surrounding code to judge it. Redaction is routinely applied in one layer and undone in another, so do not conclude from one file.
BrowserMesh drives real browsers on a user's machine. It holds cookies, storage state, credentials typed into forms, and whole page contents by design. Every one of those is one careless template literal away from stderr or an MCP result.
What must never escape
Into logs, MCP results, error messages, or diagnostics:
- cookies and auth tokens;
- saved browser storage state;
- passwords and form values;
- full page contents;
- arbitrary filesystem paths, including absolute paths in stack traces and configuration errors.
Screenshots stay in-memory MCP image results. .browsermesh/ is never committed.
Where it leaks
- Error mapping. A raw Playwright error carries selectors, URLs, and sometimes page text. Check
that concrete errors are mapped to stable
BrowserMeshErrorcodes and that the public message does not interpolate the original. Check the fatal and CLI paths too — a configuration failure should name the variable, not print its value or a stack. - Structured logging. Check what is interpolated into log records, not just the log level.
Object spreads are the usual culprit:
{...options}quietly carries whatever was added upstream. - MCP results.
structuredContentis returned verbatim to a client. Check that a result echoes identifiers, not payloads. - Persistence. Callers supply logical
stateIdvalues, never paths. Check that traversal is rejected, that a caller-controlled path cannot reach the filesystem layer, and that state contents are never logged. - Diagnostics.
--doctorand runtime info expose configuration; check that only non-sensitive effective values are reported. - stdout discipline. stdout carries MCP protocol traffic only. Any human or debug output on stdout corrupts the protocol; structured logs go to stderr.
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.
- 5d ago First seen · 62 lines · 66 tokens per session scan A 8fb825826a0a
log-leak-reviewer is an agent published in the GitHub repository scrollDynasty/BrowserMesh (4 stars, last pushed yesterday), licensed Apache-2.0. It adds 66 tokens to every session and 695 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 2026-08-31.
Other agents, from other repositories
web-research-analyst
Web research: browser automation, Tavily API, competitive intelligence, documentation capture, technical recon.
replay-agent
Build replay sequences for UI debugging, regression testing, and automation.
issue-tracker
This repository uses GitHub Issues.
claude-code-mcp-chrome-bridge-single-transport-deadlock
Agent "claude-code-mcp-chrome-bridge-single-transport-deadlock" from laozhangzzz/agent-pitbook, covering summary, symptoms, root cause, fix and verification.
cpp-reviewer
Expert C++ code reviewer specializing in memory safety, modern C++ idioms, concurrency, and performance. Use for all C++ code changes. MUST BE USED for C++ projects.
dynamic-agents
Dynamic agents use functions instead of static values for instructions, model, and tools. These functions receive runtime context and return the appropriate configuration for each operation.