Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Track finance investment signal evolution and update logic based on new finance market information. Use when monitoring finance signals and determining if they are strengthened, weakened, or falsified.
.claude/skills/rkiding-alphaear-signal-tracker/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -18% | 0% |
This skill provides logic to track and update investment signals. It assesses how new market information impacts existing signals (Strengthened, Weakened, Falsified, or Unchanged).
YOU (the Agent) are the Tracker. Use the prompts in references/PROMPTS.md.
Workflow:
InvestmentSignal.Tools:
alphaear-search and alphaear-stock skills to gather the necessary data.scripts/fin_agent.py helper _sanitize_signal_output if needing to clean JSON.Key Logic:
Example Usage (Conceptual):
python# This skill is currently a pattern extracted from FinAgent. # In a future refactor, it should be a standalone utility class. # For now, refer to `scripts/fin_agent.py`'s `track_signal` method implementation.
agno (Agent framework)sqlite3 (built-in)Ensure DatabaseManager is initialized correctly.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 15,505 | 11,122 | -28% | 1 | 1 | 0% | 2,507 | 2,215 | -12% | 0 | 0 | — |
case-02 | fail→pass | 12,354 | 6,859 | -44% | 1 | 1 | 0% | 2,100 | 1,484 | -29% | 0 | 0 | — |
case-03 | fail→pass | 13,511 | 9,237 | -32% | 1 | 1 | 0% | 2,260 | 2,065 | -9% | 0 | 0 | — |
case-04 | fail→fail | 8,028 | 5,510 | -31% | 1 | 1 | 0% | 1,324 | 1,350 | +2% | 0 | 0 | — |
case-05 | fail→pass | 9,944 | 6,900 | -31% | 1 | 1 | 0% | 1,746 | 1,663 | -5% | 0 | 0 | — |
case-06 | fail→pass | 5,247 | 1,987 | -62% | 1 | 1 | 0% | 917 | 756 | -18% | 0 | 0 | — |
case-11 | pass→pass | 3,768 | 1,524 | -60% | 1 | 1 | 0% | 590 | 597 | +1% | 0 | 0 | — |
case-07 | pass→pass | 5,483 | 1,947 | -64% | 1 | 1 | 0% | 933 | 774 | -17% | 0 | 0 | — |
case-08 | fail→pass | 9,346 | 2,125 | -77% | 1 | 1 | 0% | 1,544 | 753 | -51% | 0 | 0 | — |
case-09 | fail→pass | 6,549 | 1,435 | -78% | 1 | 1 | 0% | 1,276 | 618 | -52% | 0 | 0 | — |
case-10 | fail→pass | 10,991 | 1,385 | -87% | 1 | 1 | 0% | 1,866 | 621 | -67% | 0 | 0 | — |
case-20 | pass→pass | 11,326 | 13,213 | +17% | 1 | 1 | 0% | 2,889 | 3,780 | +31% | 0 | 0 | — |
case-12 | fail→pass | 8,201 | 1,809 | -78% | 1 | 1 | 0% | 1,345 | 718 | -47% | 0 | 0 | — |
case-13 | fail→pass | 9,533 | 1,884 | -80% | 1 | 1 | 0% | 1,605 | 705 | -56% | 0 | 0 | — |
case-14 | pass→pass | 5,277 | 1,851 | -65% | 1 | 1 | 0% | 1,063 | 691 | -35% | 0 | 0 | — |
case-15 | fail→pass | 6,195 | 1,508 | -76% | 1 | 1 | 0% | 1,078 | 682 | -37% | 0 | 0 | — |
case-16 | pass→pass | 9,383 | 5,902 | -37% | 1 | 1 | 0% | 1,538 | 1,537 | -0% | 0 | 0 | — |
case-17 | pass→pass | 13,914 | 7,485 | -46% | 1 | 1 | 0% | 2,110 | 1,703 | -19% | 0 | 0 | — |
case-18 | fail→pass | 11,253 | 7,960 | -29% | 1 | 1 | 0% | 1,930 | 1,811 | -6% | 0 | 0 | — |
case-19 | fail→pass | 7,323 | 1,328 | -82% | 1 | 1 | 0% | 1,223 | 629 | -49% | 0 | 0 | — |
case-21 | pass→pass | 12,238 | 11,392 | -7% | 1 | 1 | 0% | 2,275 | 2,733 | +20% | 0 | 0 | — |
case-22 | pass→pass | 13,636 | 12,615 | -7% | 1 | 1 | 0% | 2,865 | 2,996 | +5% | 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. The headline lift of +59 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.