fix: scope benchmark skill to web performance

[creatornader]

Summary

• rename the web performance benchmark skill to gstack-benchmark
• narrow benchmark triggers so bare benchmark requests do not match this skill
• add generator coverage for benchmark scope and valid llms.txt skill links

Tests

• bun test test/gen-skill-docs.test.ts test/llms-txt-shape.test.ts

[trunk-io]

Merging to main in this repository is managed by Trunk.

• To merge this pull request, check the box to the left or comment /trunk merge below.

After your PR is submitted to the merge queue, this comment will be automatically updated with its status. If the PR fails, failure details will also be posted here

[time-attack]

Thanks for this. The command rename leaves active references, lacks a compatibility plan, and combines an independent link-generation repair with incomplete routing coverage. Please rebase, choose one canonical command, preserve compatibility, separate the link fix, update all references, and add real routing tests.

[creatornader]

Thanks. I am addressing this review as three focused PRs:

• fix: scope benchmark skill to web performance #2083, this PR: the canonical /web-performance-benchmark command, compatibility aliases, active references, and routing coverage.
• fix: preserve generated skill links #2279: the independent generated gstack/llms.txt link repair and its link-validity coverage.
• docs: add stock phrase guardrail #1795: the unrelated voice-guardrail feedback and regenerated outputs.

The rename exposes the generator's invalid path assumption, so #2279 needs to land first. I will then rebase #2083 onto main and remove the copied link-repair diff from this branch. That leaves this PR limited to the command migration and its compatibility behavior.

The canonical command is /web-performance-benchmark. /benchmark and /gstack-benchmark remain compatibility aliases. I have updated the remaining active browser-design reference and added regression coverage for the canonical route. I will follow up after the rebase with the final verification rather than marking this feedback resolved early.

[creatornader]

Update: the two live routing selections now pass locally. A performance-regression request invokes web-performance-benchmark, and an explicit /benchmark request resolves to the same canonical workflow.

The remaining dependency is #2279. Trunk rejected my submission from this fork because I do not have write access to the repository. Once a maintainer submits and lands #2279, I will rebase this PR onto main, remove the duplicated link repair, and post the final focused diff for review.