Koen Vermeulen
Editor tooling and MCP wiring.
Member · 309 reputation · here since 27 May 2025
Asked (14)
Can --explain show the values a layer lost with, not just the winner?
The provenance table gives one row per leaf with the winning layer. Passing --headed shows runtime.headed cli; exporting SDODS_HEADED=1 instead shows processEnv; with both it…
asked by Koen Vermeulen309 ·Is there a flag to raise the analyze file budget? --max-files is not accepted
The truncation warning tells me the scan hit its limit. Nothing in the analyze reference looks like a budget flag, and guessing does not help: Same for --max-depth. Am I missing…
asked by Koen Vermeulen309 ·Every scenario passed and the release gate still failed the job
Nothing failed. Zero. And the job is red, so the release is blocked and I am the one holding it. I can see the two flaky scenarios in the report and they both ended up passing.…
asked by Koen Vermeulen309 ·Docs call the insights CLI a placeholder but it clearly does something
The known limitations table says: It does not behave like a placeholder: wrote .sdods/insights/demo-shop.json, printed suite health with a trend, a scenarios-by-flakiness table…
asked by Koen Vermeulen309 ·Can I keep two SDODS installs on one machine without them fighting?
I need to reproduce a run from a while back against the build the team was standardised on then, while keeping my normal install for day-to-day work. Is that supported or am I…
asked by Koen Vermeulen309 ·Why does the HAR strict failure have no code, hint or docs link like every other error?
Small thing, but it broke a script. Every failure I have seen so far comes out shaped like this, and our CI wrapper parses the code out of it: A strict HAR miss comes out shaped…
asked by Koen Vermeulen309 ·record convert: Cannot tell which project my spec belongs to, and I only have one project
Copied a recording out to a scratch directory to try the conversion without touching the repo: There is one project in the whole workspace. Why does it not just pick it?
asked by Koen Vermeulen309 ·Same locator healed five times and I still have no page-object proposal
The docs say three identical successful heals of the same locator produce a page-object patch proposal. Ours has healed five times across four nightly runs, same description,…
asked by Koen Vermeulen309 ·Two Examples blocks, one @smoke and one @regression — lint says found @smoke @regression
I wanted one outline with a small set of rows in smoke and the long tail in regression: Both tables are tagged once. Is this rule just not aware of Examples-level tags?
asked by Koen Vermeulen309 ·products.json must contain an array of rows (or { rows: [...] }) — mine is keyed by sku
The export from our catalogue service is an object keyed by sku, which is the shape everything downstream of it expects: Pointing a dataset at it gives: Odd part: sdods data…
asked by Koen Vermeulen309 ·Tagged a scenario @a11y and it went green in four seconds — how do I tell if axe ran?
Added @a11y to two existing scenarios expecting them to get slower and probably to fail, because the pages in question are not good. Both passed, in about the same time as…
asked by Koen Vermeulen309 ·Three projects with nearly identical settings — how much can actually be shared?
Our three project files are 80% the same: same browsers, same suites, same four processes, same timeouts. Every change is a three-file change and one of them always gets missed.…
asked by Koen Vermeulen309 ·The reference says 37 tools, the MCP guide says 33 to 35 — which number is right?
Writing an internal page about what our agents can reach and I cannot make the docs agree with themselves. The MCP tools reference is titled as the 37 tools. The MCP guide says…
asked by Koen Vermeulen309 ·Is XPath actually forbidden in a page object, or only discouraged?
Reviewer left "never XPath" on my proposal. The locator strategy page says CSS or XPath as a last resort inside a page object with a heal context, which reads like "allowed but…
asked by Koen Vermeulen309 ·
Answered (7)
Local model loops on agent generate, calling the same tool until it hits max turns
It calls feature_list, reads the result, calls feature_list again with the same arguments, and keeps going until the turn budget stops it. No proposal. It is not confused about…
asked by Paulo Mendes97 ·Should each tester run sdods mcp locally, or should we host one server for the team?
Eight testers, one repo, one staging environment. Right now everybody has stdio in their editor and their own checkout. It works, but nobody can ask about a run somebody else…
asked by Ingrid Solberg241 ·proposals accept --branch fails with Could not switch to branch and no reason why
That is the whole message. Nothing applied, which I am glad about, but git's own reason is nowhere — not on stdout, not with --json. Am I supposed to guess?
asked by Devon Marsh183 ·sdods mcp install claude dies with claude mcp add failed and writes nothing
On a locked-down work laptop: I do not have write access to that file and IT will not change it this quarter. Does the hint actually help me, or is it just a different way to fail?
asked by Helle Borg352 ·How do I give the team coding agents read-only access to our SDODS server?
We host one SDODS server for the QA group. Everyone wants to point their editor at /mcp so they can ask about runs and coverage without a checkout. What I do not want is…
asked by Rasha Halabi572 ·My mcp.servers block is ignored when the agent runs with the openai adapter
Project yaml has the usual: With --adapter claude the planner browses the app and the plan is full of real page names. With --adapter openai against our internal endpoint, the…
asked by Thom Vasseur5,470 ·sdods auth capture --index 3 says Role "admin" has 2 user(s); index 3 does not exist
The CSV has six rows and the admin I want is on the fourth one, so index 3 by my counting. Where does "2 user(s)" come from?
asked by Tanvi Bhatt126 ·