Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Resolves a PostHog experiment reference from natural language to a concrete experiment ID by browsing `experiment-list` (not feature-flag tools), with disambiguation when multiple experiments match. Use when the user names or quotes an experiment ("split test demo", "the File engagement boost experiment", "onboarding retention test", "landing page hero experiment", "pricing experiment"), describes it loosely ("the signup experiment", "my pricing test", "the one with the new checkout"), uses a re
.claude/skills/kunanonj-cursor-plugin-posthog-finding-experiments/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 256% | 0% |
| case-05 | ✓→✗ | ▼ Worse | -28% | 0% |
Users refer to experiments by name, description, or relative references — not by ID. This skill resolves natural language references to concrete experiment IDs.
Use the experiment-list tool from the Posthog-local MCP server.
IMPORTANT: Do NOT use feature-flag-get-all or any feature flag tool to find experiments. Use the dedicated experiment list tool: experiment-list.
This tool returns experiments with their id, name, status, feature_flag_key, start_date, end_date, and created_at. Browse the returned list to find the experiment matching the user's reference:
name field for matchesstatus field (draft, running, stopped)feature_flag_key fieldAfter resolving to an ID, call experiment-get for the full object (metrics, flag details, parameters).
textUser: "pause my signup experiment" Agent: 1. Calls experiment-list 2. Scans results, finds "New signup process" (ID: 1371, status: running) 3. Proceeds to pause experiment 1371
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→fail | 6,618 | 6,511 | -2% | 1 | 1 | 0% | 1,062 | 764 | -28% | 0 | 0 | — |
case-01 | fail→fail | 2,386 | 2,501 | +5% | 1 | 1 | 0% | 455 | 615 | +35% | 0 | 0 | — |
case-02 | fail→fail | 4,459 | 2,538 | -43% | 1 | 1 | 0% | 368 | 613 | +67% | 0 | 0 | — |
case-03 | fail→fail | 3,592 | 2,201 | -39% | 1 | 1 | 0% | 327 | 589 | +80% | 0 | 0 | — |
case-04 | fail→fail | 5,084 | 2,590 | -49% | 1 | 1 | 0% | 1,031 | 578 | -44% | 0 | 0 | — |
case-06 | fail→fail | 5,017 | 4,784 | -5% | 1 | 1 | 0% | 777 | 733 | -6% | 0 | 0 | — |
case-07 | fail→fail | 7,342 | 2,704 | -63% | 1 | 1 | 0% | 1,280 | 593 | -54% | 0 | 0 | — |
case-08 | fail→pass | 9,232 | 4,439 | -52% | 1 | 1 | 0% | 1,610 | 1,128 | -30% | 0 | 0 | — |
case-09 | fail→pass | 10,766 | 3,436 | -68% | 1 | 1 | 0% | 1,348 | 916 | -32% | 0 | 0 | — |
case-10 | fail→pass | 5,517 | 3,484 | -37% | 1 | 1 | 0% | 970 | 1,115 | +15% | 0 | 0 | — |
case-11 | fail→fail | 7,270 | 2,057 | -72% | 1 | 1 | 0% | 1,255 | 518 | -59% | 0 | 0 | — |
case-12 | fail→fail | 8,924 | 6,475 | -27% | 1 | 1 | 0% | 1,575 | 648 | -59% | 0 | 0 | — |
case-13 | pass→fail | 6,908 | 3,214 | -53% | 1 | 1 | 0% | 1,345 | 610 | -55% | 0 | 0 | — |
case-14 | fail→fail | 7,164 | 2,592 | -64% | 1 | 1 | 0% | 1,178 | 565 | -52% | 0 | 0 | — |
case-15 | fail→fail | 6,295 | 2,771 | -56% | 1 | 1 | 0% | 999 | 555 | -44% | 0 | 0 | — |
case-16 | fail→pass | 2,466 | 5,799 | +135% | 1 | 1 | 0% | 385 | 1,371 | +256% | 0 | 0 | — |
case-17 | fail→fail | 2,875 | 2,315 | -19% | 1 | 1 | 0% | 473 | 549 | +16% | 0 | 0 | — |
case-18 | pass→pass | 3,601 | 4,724 | +31% | 1 | 1 | 0% | 510 | 1,164 | +128% | 0 | 0 | — |
case-19 | fail→fail | 7,069 | 4,609 | -35% | 1 | 1 | 0% | 1,454 | 664 | -54% | 0 | 0 | — |
case-20 | fail→fail | 3,693 | 1,957 | -47% | 1 | 1 | 0% | 724 | 529 | -27% | 0 | 0 | — |
case-21 | pass→pass | 5,415 | 1,955 | -64% | 1 | 1 | 0% | 981 | 784 | -20% | 0 | 0 | — |
case-22 | fail→fail | 4,573 | 2,686 | -41% | 1 | 1 | 0% | 296 | 638 | +116% | 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 6 counted toward the lift figure. The other 16 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 +9 percentage points is the difference between those two pass rates over the 6 comparable cases. 6 cases got worse with the skill loaded, and they are 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.