Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Diagnose Basecamp CLI, authentication, and agent-plugin health.
.claude/skills/basecamp-basecamp-doctor/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -74% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -76% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -54% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -84% | 0% |
Run the structured diagnostic:
bashbasecamp doctor --json
Interpret every check by status:
pass: working correctly.warn: usable, but follow-up is recommended.skip: not run because it is unauthenticated or not applicable.fail: broken and needs attention.Report failures and warnings with their hint fields. Also inspect the top-level breadcrumbs array and preserve its structured cmd next steps, because a breadcrumb can provide a more specific action than a check hint. Use these common remediations when relevant:
basecamp auth login. Guard a replacement loginwith --expect-identity <id> so a browser signed in as someone else cannot become the profile. A bot or CI profile that must never sign in interactively imports a personal access token instead: op read "op://<vault>/<item>/credential" | basecamp auth login --with-token -P <profile> --account <id> --expect-identity <id>
Check oauth_type before suggesting either. basecamp auth status --json reports it, and agent means the profile is a Basecamp agent: a principal with no person behind it, which authenticates with its OAuth client rather than a sign-in. Both commands above would store a PERSON's credential under that profile and silently replace the agent. Its remediation is its own login, with the client secret piped in: op read "op://<vault>/<item>/credential" | basecamp auth login --with-client-credentials --client-id <id> -P <profile> --account <id> The CLI's own hint on an agent credential already names this command with the client id filled in — prefer it verbatim over reconstructing one, and follow it rather than choosing for yourself whenever oauth_type is absent.
basecamp setup agents (honors BASECAMP_SETUP_AGENT)basecamp setup codexbasecamp setup claudebasecamp setup grok (skill-only; Grok reads the shared ~/.agents/skills/basecamp skill)Every remediation above runs without a terminal. Bare basecamp setup is the human first-time flow and is not one of them: it opens browser OAuth and refuses with a usage error in machine-output modes or unless stdin, stdout, and stderr are all terminals. basecamp setup --customize additionally asks the user to choose each default. Suggest either to a human at a terminal if useful, but never run them yourself — use the subcommands above.
Do not read, print, or request credential files. If every check passes, say that Basecamp and its agent integration are ready.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 26,370 | 14,067 | -47% | 1 | 1 | 0% | 2,322 | 577 | -75% | 0 | 0 | — |
case-02 | fail→fail | 8,322 | 4,143 | -50% | 1 | 1 | 0% | 1,325 | 427 | -68% | 0 | 0 | — |
case-03 | fail→fail | 31,660 | 14,755 | -53% | 1 | 1 | 0% | 4,745 | 620 | -87% | 0 | 0 | — |
case-04 | fail→pass | 23,968 | 7,741 | -68% | 1 | 1 | 0% | 1,892 | 841 | -56% | 0 | 0 | — |
case-10 | pass→pass | 10,579 | 7,702 | -27% | 1 | 1 | 0% | 909 | 747 | -18% | 0 | 0 | — |
case-05 | fail→pass | 21,728 | 2,236 | -90% | 1 | 1 | 0% | 2,670 | 693 | -74% | 0 | 0 | — |
case-06 | fail→pass | 21,079 | 7,116 | -66% | 1 | 1 | 0% | 2,714 | 658 | -76% | 0 | 0 | — |
case-07 | pass→pass | 11,681 | 3,795 | -68% | 1 | 1 | 0% | 1,756 | 973 | -45% | 0 | 0 | — |
case-08 | pass→pass | 17,021 | 8,803 | -48% | 1 | 1 | 0% | 1,846 | 959 | -48% | 0 | 0 | — |
case-09 | pass→pass | 15,245 | 8,840 | -42% | 1 | 1 | 0% | 1,821 | 771 | -58% | 0 | 0 | — |
case-11 | fail→pass | 7,609 | 6,473 | -15% | 1 | 1 | 0% | 1,137 | 525 | -54% | 0 | 0 | — |
case-12 | pass→pass | 12,658 | 9,873 | -22% | 1 | 1 | 0% | 2,034 | 1,150 | -43% | 0 | 0 | — |
case-13 | pass→pass | 7,904 | 7,659 | -3% | 1 | 1 | 0% | 1,174 | 754 | -36% | 0 | 0 | — |
case-14 | pass→pass | 7,470 | 8,220 | +10% | 1 | 1 | 0% | 412 | 844 | +105% | 0 | 0 | — |
case-15 | fail→pass | 39,296 | 8,478 | -78% | 1 | 1 | 0% | 5,627 | 877 | -84% | 0 | 0 | — |
case-16 | fail→pass | 9,752 | 4,879 | -50% | 1 | 1 | 0% | 1,586 | 1,136 | -28% | 0 | 0 | — |
case-17 | fail→pass | 18,884 | 7,303 | -61% | 1 | 1 | 0% | 2,132 | 699 | -67% | 0 | 0 | — |
case-18 | pass→pass | 8,957 | 3,839 | -57% | 1 | 1 | 0% | 1,253 | 662 | -47% | 0 | 0 | — |
case-19 | pass→pass | 15,661 | 2,765 | -82% | 1 | 1 | 0% | 1,714 | 752 | -56% | 0 | 0 | — |
case-20 | pass→fail | 11,377 | 3,709 | -67% | 1 | 1 | 0% | 2,115 | 446 | -79% | 0 | 0 | — |
case-21 | pass→pass | 13,071 | 27,257 | +109% | 1 | 1 | 0% | 2,562 | 3,702 | +44% | 0 | 0 | — |
case-22 | pass→fail | 10,563 | 9,266 | -12% | 1 | 1 | 0% | 1,914 | 449 | -77% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 17 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +23 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
The publisher has shipped newer versions since this run, so these numbers describe v3, not the version currently listed.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/27/2026 | +41% |
| gemini-3.6-flash | verified | 8/24/2026 | +55% |
Other measured skills in the registry, with their headline benchmark lift.