pnpm lint causing issue when ESLint isn't used

[timothyclin]

I have a project that uses pnpm and biome as linter.
RTK seems to insist rewrite pnpm lint to rtk lint, which is then dependent on ESLint that I do not use.
This causes the agent to get into a re-try loop. Despite several notes I provided in Agents.MD, the confusion seems big enough that the agents kept tripping over this. Sometimes it'd eventually figure out what to do.

Proof -

# Test RTK rewrite of pnpm lint
$ rtk rewrite "pnpm lint"
rtk lint

[timothyclin]

my current workaround is to use pnpm run lint, which rtk doesn't rewrite.

[TechWizard9999]

I’m taking this issue. I’ll adjust rewrite behavior so pnpm lint is routed through rtk pnpm lint (not rtk lint) to avoid ESLint-specific loops, with tests, in a focused PR against develop.

[TechWizard9999]

I’ve submitted a PR to address this issue: #678

Changes include:

• routing pnpm lint to rtk pnpm lint instead of forcing rtk lint
• preserving pnpm lint arguments in rewrite output
• adding regression tests for both plain and argumented pnpm lint rewrites

Validation run:

• cargo fmt --all --check
• rtk cargo clippy --all-targets
• rtk cargo test
• ./target/debug/rtk rewrite "pnpm lint" => rtk pnpm lint

Please review when you get a chance.