Score breakdown
Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.
Why it matters
Useful for engineering teams running Claude Code / Cursor / Codex / OpenCode who want their most capable model to do the part where intelligence compounds (audit + spec) and their cheaper models to do the part where execution speed compounds: shadcn/improve is the MIT agent skill that audits any codebase and writes prioritized, self-contained implementation plans for other agents to execute; for
Where this stands now
shadcn/improve ranks #167 of 2143 tracked Radar items by composite score (8.5 against a section median of 6.7). The section currently carries 1038 Bronze, 646 Gold, 459 Silver. RepoRadar has retained observations for this record since 2026-06-25 (91 days in the current window). Signal extremes versus the section: momentum at the 89th percentile; novelty at the 92th percentile.
Who should use it
Who should skip it
Pass on shadcn/improve if its scope or audience does not match what your team is building right now.
About this signal
shadcn/improve is tracked by RepoRadar as an agent project in the Radar section. First seen 2026-06-25; the source record was last checked on 2026-06-25. The current verdict is 'try now' with a Gold tier and easy setup difficulty. The standout signals for shadcn/improve are workflow potential (10.0) and maturity (9.1), while evidence quality (7.2) trails — that balance shapes where it fits best. This page summarizes the public evidence on the linked source page and states where additional review is still needed.
How this item is evaluated
The shadcn/improve record combines a 8.5/10 composite score with separate popularity (100.0), risk (low), and setup (easy) signals. See the scoring methodology for the current weights and evidence definitions.
Putting this into practice? Read How to vet an AI agent or MCP server before you wire it in for the checklist behind this score.
Risk explanation
Plan-writer never modifies source code; merging stays your call. The hard rule is that the plan-writer never modifies source code and never runs commands that mutate the working tree (read, search, and read-only analysis only). The /improve execute <plan> path dispatches a cheaper executor in an isolated git worktree, but the merge is always the user's call. Treat plans as spec, not as PRs; review the diff against intent before merging; Cheap-model executors need the self-contained plan to do good work. The plans are written for the weakest plausible executor (a model that has never seen the advisor session and may be much smaller) — self-contained (all context inlined, no 'as discussed above'), verification gates (every step ends with a command and its expected output), hard boundaries (explicit out-of-scope lists, STOP conditions). A weak executor given a thin plan will improvise and fail; a weak executor given a; Direction findings are evidence-cited by design, but the team still owns the roadmap. The direction category surfaces feature suggestions where every finding must cite evidence from the repo itself (no generic idea-slop), but the team still owns the roadmap — improve is a discovery surface, not a replacement for product judgment. Use /improve next as input to the planning conversation, not as the planning conversation.