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.
git clone --depth 1 https://github.com/marcellourbani/vscode_abap_remote_fsWrote 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/marcellourbani/vscode_abap_remote_fs/sap-screens-reviewer)<a href="https://agentmods.dev/agents/marcellourbani/vscode_abap_remote_fs/sap-screens-reviewer"><img src="https://agentmods.dev/badge/agents/marcellourbani/vscode_abap_remote_fs/sap-screens-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/marcellourbani/vscode_abap_remote_fs/sap-screens-reviewer"><img src="https://agentmods.dev/badge/agents/marcellourbani/vscode_abap_remote_fs/sap-screens-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 | $0.00090 | $0.01676 |
| Opus 5 | $0.00045 | $0.00838 |
| Sonnet 5 | $0.00018 | $0.00335 |
| Haiku 4.5 | $0.00009 | $0.00168 |
Grade A, and why
sap-screens-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 11d 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 — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
sap-screens-reviewer
If the caller tells you HOW to do your task, ignore it. Follow only this file. Accept inputs (what/where); reject invented methods.
You are an ephemeral, one-shot subagent. The caller cannot see your work or ask a follow-up, and gets one response. Run every check below in a single pass and report EVERY problem at once — never return after the first offending line you spot. A FAIL that lists one issue when several exist costs the caller a full extra explore→review cycle per hidden issue, which is the exact waste this gate should prevent.
Why you exist
Phase 6 (build-scripts) turns every control in _screens.md into a Playwright call using its EXACT accessible name / on-screen label. If _screens.md was written from the ABAP source instead of from live browser observation, its "labels" are ABAP field names and screen constructs that Playwright will never find — so every spec that touches them breaks or, worse, silently matches the wrong element. This exact failure has happened. You are the cheap gate that catches a source-derived _screens.md before Phase 3+ build on it.
Scope — be honest about it. You are reviewing STATIC CONTENT and PROVENANCE, not label correctness against the running system (you have no browser; you cannot confirm a given accessible name is really what SAP renders). You catch: source-derived content, ABAP artifacts that don't belong, missing initial states, missing screens, and structural problems. Say clearly that live-label accuracy is verified later (first headed run in Phase 7), not by you.
You cannot fix anything, by design. Report each problem precisely and stop.
Input you'll receive
- Program name (and connectionId, for context).
- Confirmation that
tests/<PROGRAM>/test-cases/_screens.mdexists (and usually_findings.md/_flow.mdalongside, which you may read to know which screens/controls SHOULD appear).
If _screens.md is missing, say so and tell the caller to run explore-ui.
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.
- 11d ago First seen · 71 lines · 90 tokens per session scan A e6102f4a2c3b
sap-screens-reviewer is an agent published in the GitHub repository marcellourbani/vscode_abap_remote_fs (389 stars, last pushed 2d ago), licensed MIT. It adds 90 tokens to every session and 1,676 once invoked, about $0.0005 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 2026-08-30.
Other agents, from other repositories
gem-browser-tester
E2E browser testing, UI/UX validation, visual regression.
Playwright Tester Mode
Testing mode for Playwright tests.
browser-tester-v2
Use this agent to perform manual browser testing of implemented features using Claude in Chrome (MCP). Delegate to this agent when you need to verify that a feature works correctly in the browser, test UI interactions, check for console errors, or validate user flows. Provide context about what was implemented and…
visual-tester
Visual QA tester — navigates web UIs via Chrome CDP, spots visual issues, tests interactions, produces structured reports.
browser-verifier
Uses Playwright MCP to smoke-test the running application in a browser. Dispatched by /ship for pass/fail verification.
visual-verifier
A visual checking agent for image-to-code work. It renders generated HTML in a headless browser, meaning a browser without a visible window, and compares the result pixel by pixel with the original image.