Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Look up and update Asana tasks using local developer-workflow config and direct Asana REST calls. Use when the user asks to sync an Asana task, move task status, add a PR link to Asana, or validate Asana configuration.
.claude/skills/tetherto-qv-asana-sync/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 57% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 48% | 0% |
Manual Asana task lookup and lifecycle sync using direct REST helpers. No Asana MCP binary.
If the user invokes this skill without a ticket, task GID, or specific action, do not only run config validation and exit. Show the operation menu below and ask what they want to do.
Ask for the missing input:
find, status, move, comment, or completemovecomment~/.config/qvac-pr-skills/config.jsonASANA_ACCESS_TOKEN set in the current environment or available from the user's default shell startup files, unless config uses a different asana.tokenEnvCreate the token:
Cursor local qvac-pr-skills.bashexport ASANA_ACCESS_TOKEN='<token>'
bashcurl https://app.asana.com/api/1.0/users/me \ -H "Authorization: Bearer $ASANA_ACCESS_TOKEN"
Docs: Asana personal access token
Validate:
bashnode .cursor/skills/_lib/developer-workflow/config-init.mjs --validate --asana
Show these options when no action is provided:
status - validate local Asana config and token accessfind <ticket> - find an Asana task by tickettask <task-gid> - read a task by Asana GIDmove <ticket-or-task-gid> <status> - move task to configured statuscomment <ticket-or-task-gid> <text> - add task commentcomplete <ticket-or-task-gid> - mark task complete and move to completedAvailable statuses: todo, inProgress, blocked, inReview, readyForQa, qaPassed, staging, storeReview, readyForProd, completed, closed.
Status / config validation:
bashnode .cursor/skills/_lib/developer-workflow/asana-cli.mjs status
Find a task by ticket:
bashnode .cursor/skills/_lib/developer-workflow/asana-cli.mjs find QVAC-12345
Read a task by Asana task GID:
bashnode .cursor/skills/_lib/developer-workflow/asana-cli.mjs task <task-gid>
Move a ticket/task to a configured status. If asana.statusOptions is configured, this uses Asana's custom_type_status_option. Otherwise it falls back to old section movement:
bashnode .cursor/skills/_lib/developer-workflow/asana-cli.mjs move QVAC-12345 inProgress node .cursor/skills/_lib/developer-workflow/asana-cli.mjs move QVAC-12345 inReview node .cursor/skills/_lib/developer-workflow/asana-cli.mjs move QVAC-12345 blocked node .cursor/skills/_lib/developer-workflow/asana-cli.mjs move QVAC-12345 completed
Add a comment, for example a PR URL:
bashnode .cursor/skills/_lib/developer-workflow/asana-cli.mjs comment QVAC-12345 "PR opened: https://github.com/owner/repo/pull/123"
Complete a task and move it to configured completed status:
bashnode .cursor/skills/_lib/developer-workflow/asana-cli.mjs complete QVAC-12345
Use this mapping after user confirmation:
move <ticket> inProgressmove <ticket> inReview, then comment <ticket> "PR opened: <url>"move <ticket> blocked, then comment with blocker contextcomplete <ticket>For mutations, show the exact planned command first and wait for explicit user confirmation.
/users/me to discover the current user; never copy another developer's user GID.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-18 | pass→pass | 10,379 | 2,770 | -73% | 1 | 1 | 0% | 1,634 | 1,708 | +5% | 0 | 0 | — |
case-01 | fail→pass | 12,583 | 3,808 | -70% | 1 | 1 | 0% | 1,274 | 2,000 | +57% | 0 | 0 | — |
case-02 | fail→pass | 14,440 | 4,559 | -68% | 1 | 1 | 0% | 2,750 | 2,033 | -26% | 0 | 0 | — |
case-03 | fail→fail | 6,814 | 8,401 | +23% | 1 | 1 | 0% | 381 | 1,658 | +335% | 0 | 0 | — |
case-04 | fail→pass | 8,364 | 2,470 | -70% | 1 | 1 | 0% | 1,207 | 1,597 | +32% | 0 | 0 | — |
case-05 | fail→pass | 10,310 | 2,780 | -73% | 1 | 1 | 0% | 1,812 | 1,683 | -7% | 0 | 0 | — |
case-06 | fail→pass | 6,522 | 1,955 | -70% | 1 | 1 | 0% | 1,017 | 1,510 | +48% | 0 | 0 | — |
case-07 | fail→pass | 2,862 | 2,306 | -19% | 1 | 1 | 0% | 423 | 1,603 | +279% | 0 | 0 | — |
case-08 | fail→pass | 10,240 | 2,762 | -73% | 1 | 1 | 0% | 1,793 | 1,627 | -9% | 0 | 0 | — |
case-09 | fail→pass | 9,189 | 3,426 | -63% | 1 | 1 | 0% | 1,436 | 1,800 | +25% | 0 | 0 | — |
case-10 | fail→fail | 12,741 | 6,940 | -46% | 1 | 1 | 0% | 2,073 | 1,700 | -18% | 0 | 0 | — |
case-11 | fail→pass | 7,742 | 2,848 | -63% | 1 | 1 | 0% | 1,238 | 1,679 | +36% | 0 | 0 | — |
case-12 | fail→pass | 6,979 | 2,461 | -65% | 1 | 1 | 0% | 1,300 | 1,629 | +25% | 0 | 0 | — |
case-13 | pass→pass | 9,325 | 4,036 | -57% | 1 | 1 | 0% | 1,646 | 1,938 | +18% | 0 | 0 | — |
case-19 | pass→pass | 3,896 | 3,328 | -15% | 1 | 1 | 0% | 582 | 1,788 | +207% | 0 | 0 | — |
case-14 | pass→pass | 10,836 | 4,871 | -55% | 1 | 1 | 0% | 1,835 | 2,041 | +11% | 0 | 0 | — |
case-15 | pass→pass | 6,619 | 2,640 | -60% | 1 | 1 | 0% | 1,128 | 1,608 | +43% | 0 | 0 | — |
case-16 | pass→pass | 4,124 | 1,557 | -62% | 1 | 1 | 0% | 581 | 1,406 | +142% | 0 | 0 | — |
case-17 | fail→pass | 4,335 | 1,512 | -65% | 1 | 1 | 0% | 638 | 1,377 | +116% | 0 | 0 | — |
case-20 | fail→fail | 9,035 | 3,315 | -63% | 1 | 1 | 0% | 1,729 | 1,798 | +4% | 0 | 0 | — |
case-21 | pass→fail | 6,802 | 6,466 | -5% | 1 | 1 | 0% | 1,225 | 2,261 | +85% | 0 | 0 | — |
case-22 | pass→pass | 7,576 | 5,867 | -23% | 1 | 1 | 0% | 1,463 | 2,107 | +44% | 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 21 counted toward the lift figure. The other 1 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 +33 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
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.
Other measured skills in the registry, with their headline benchmark lift.