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 ss26367098/rawtrace-mcp --skill rawtrace-debug-browsergit clone --depth 1 https://github.com/ss26367098/rawtrace-mcpWrote 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/ss26367098/rawtrace-mcp/rawtrace-debug-browser)<a href="https://agentmods.dev/skills/ss26367098/rawtrace-mcp/rawtrace-debug-browser"><img src="https://agentmods.dev/badge/skills/ss26367098/rawtrace-mcp/rawtrace-debug-browser/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/ss26367098/rawtrace-mcp/rawtrace-debug-browser"><img src="https://agentmods.dev/badge/skills/ss26367098/rawtrace-mcp/rawtrace-debug-browser.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 | $0.00056 | $0.00550 |
| Opus 5.5 | $0.00022 | $0.00220 |
| Sonnet 5.5 | $0.00011 | $0.00110 |
| Haiku 4.5 | $0.00006 | $0.00055 |
Grade A, and why
rawtrace-debug-browser 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 17d 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 — 33 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Diagnose browser transitions with RawTrace
RawTrace 0.4 is a passive recorder with nine capture/query tools. Browser operation belongs to the user's existing controller; no browser action, screenshot, eval, credential-editing or tool-profile interface exists here.
Capture an authorized reproduction
Use the CDP endpoint supplied by the user or established by their authorized development workflow. If it is unknown, obtain it from the controller setup; do not invent an endpoint or attach an unrelated browser.
- Call capture_targets with cdpUrl and acknowledgeRawCapture:true; select the relevant real target IDs.
- Call capture_start with the endpoint, targetIds and acknowledgeRawCapture:true before reproducing the issue. Retain sessionId and selected target IDs.
- Reproduce through the existing browser controller or let the user reproduce manually. RawTrace does not capture that controller's action log; correlate only evidence actually recorded.
- Stop with capture_stop when the intended observation is complete, unless the user requested continued recording.
- Read trace_info and narrowly filtered trace_events. Use afterSeq/nextAfterSeq for pagination and searchBodies:true only when network body text is relevant.
- Use trace_state with pageId equal to a selected targetId and either atSeq or atTime to compare recorded states. Inspect complete/gaps before drawing conclusions.
- Read referenced large results with trace_artifact, following nextOffset until hasMore is false. Use base64 for byte-exact assembly.
For already authorized development/test capture, supply the required acknowledgment without asking for duplicate consent. Ask only when the actual target/account/data authorization is unclear. This workflow does not expand the authorization of other browser tools.
All capture categories default to enabled. Disable categories the user excludes; Cookie scope is the selected pages' shared browser contexts, not individual tabs. Incremental storage does not redact passwords, cookies, tokens or personal data.
What ships with it
1 file 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.
- 17d ago Changed · -32 lines · -21 tokens per session f853be082aa5
- 1mo ago First seen · 65 lines · 77 tokens per session scan A c42d78835b69
rawtrace-debug-browser is a skill published in the GitHub repository ss26367098/rawtrace-mcp (1 stars, last pushed 18d ago), licensed MIT. It adds 56 tokens to every session and 550 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
moli-webfetch
Fetch, inspect, crawl, and capture live, JavaScript-rendered websites with Moli. Use when Codex needs current web content, web research, fact lookup, link following, a bounded crawl, client-rendered or response-gated content, network diagnostics, or a standalone HTML, Markdown, JSON, semantic-tree, viewport or…
moli-cdp-server
Start Moli's CDP server and connect Playwright, Puppeteer, or raw CDP clients. Use to run a headless-browser CDP endpoint, replace a Chromium process, attach over CDP, enable real layout and screenshot surfaces, or diagnose CDP discovery, connection, and target startup—even when Moli is not named.
praman-sap-cli
SAP UI5 test automation via Playwright CLI. Use when testing SAP Fiori apps, discovering UI5 controls, debugging Praman tests, or automating SAP workflows. Extends playwright-cli with SAP/UI5 awareness.
CBrowser
Cognitive Browser — AI browser automation that simulates real user cognition. 122 MCP tools, 21 personas, 25 cognitive traits, constitutional safety, cognitive journeys with LLM-free trace replay, screen capture to GIF/WebP/WebM with an AI-readable frame manifest (v18.69.3). USE WHEN browser automation OR cognitive…
gem-pw
Browser-automated Gemini Gem CLI (CDP-first) — chat, CRUD, image gen, deep research, --collab surgical-diff review. No API key.
playwright-e2e-testing
Production-grade Playwright end-to-end testing skill for AI coding agents. Provides specialized guidance for writing, debugging, and maintaining Playwright tests in TypeScript, JavaScript, and Python. Covers the full testing lifecycle: test structure and architecture (Page Object Model, fixtures, custom matchers)…