main
md 66 lines 4.09 KB
Rendered Raw
1 # Prediction Scorecard: 2026-W23
2
3 - Date: 2026-06-01
4 - Validation window: up to 4 weeks later, using the furthest raw week available
5 - Predictions found: 27
6 - Predictions validated: 0
7 - Overall accuracy: 0% (0/0 correct)
8
9 ## Prediction Registry Format
10
11 Use `predictions: [{repo, direction, confidence}]` in analysis frontmatter. Directions are `up`, `flat`, or `down`; the validator will still infer Signal/Noise/Gaps from article sections when the registry is absent.
12
13 ## Quality Trend
14
15 - Summaries tracked: 3
16 - Average quality score: 74.3
17 - Trend: stable
18 - Latest week: 2026-W23 (73)
19 - Best week: 2026-W21 (76)
20 - Lowest week: 2026-W23 (73)
21
22 ## Accuracy by Claim
23
24 No validated claim calls yet.
25
26 ## Accuracy by Direction
27
28 No validated direction calls yet.
29
30 ## Validated Calls
31
32 No predictions have enough later data to score yet.
33
34 ## Insufficient Evidence
35
36 - `2026-W22` `perplexityai/bumblebee` (signal/up) — No later raw week is available inside the validation window.
37 - `2026-W22` `apple/corecrypto` (signal/up) — No later raw week is available inside the validation window.
38 - `2026-W22` `open-gsd/get-shit-done-redux` (signal/up) — No later raw week is available inside the validation window.
39 - `2026-W22` `aref-vc/tufte-claude-skill` (signal/up) — No later raw week is available inside the validation window.
40 - `2026-W22` `kingbootoshi/directional-prompting` (signal/up) — No later raw week is available inside the validation window.
41 - `2026-W22` `akitaonrails/ai-memory` (signal/up) — No later raw week is available inside the validation window.
42 - `2026-W22` `NanoFlow-io/engram` (signal/up) — No later raw week is available inside the validation window.
43 - `2026-W22` `jianshuo/ccglass` (signal/up) — No later raw week is available inside the validation window.
44 - `2026-W22` `0xSero/codex-shim` (signal/up) — No later raw week is available inside the validation window.
45 - `2026-W22` `workos/auth.md` (signal/up) — No later raw week is available inside the validation window.
46 - `2026-W22` `MoonshotAI/kimi-code` (signal/up) — No later raw week is available inside the validation window.
47 - `2026-W22` `nkzw-tech/cloudsail` (gap/up) — No later raw week is available inside the validation window.
48 - `2026-W22` `scheidydude/codeindex` (gap/up) — No later raw week is available inside the validation window.
49 - `2026-W23` `op7418/guizang-social-card-skill` (signal/up) — No raw payload exists for the prediction week.
50 - `2026-W23` `helloianneo/ian-xiaohei-illustrations` (signal/up) — No raw payload exists for the prediction week.
51 - `2026-W23` `nekocode/filetree-skill` (signal/up) — No raw payload exists for the prediction week.
52 - `2026-W23` `Christian-Katzmann/app-it` (signal/up) — No raw payload exists for the prediction week.
53 - `2026-W23` `garnix-io/garnix-ci` (signal/up) — No raw payload exists for the prediction week.
54 - `2026-W23` `qianzii2/rockduck` (signal/up) — No raw payload exists for the prediction week.
55 - `2026-W23` `QwenLM/Qwen-VLA` (signal/up) — No raw payload exists for the prediction week.
56 - `2026-W23` `Signal-Trade-Core/weather-prediction-bot` (noise/flat) — No raw payload exists for the prediction week.
57 - `2026-W23` `Trade-Execution-Labs/polymarket-sports-trading-bot` (noise/flat) — No raw payload exists for the prediction week.
58 - `2026-W23` `polymaxi2/polymarket-arbitrage-trading-bot` (noise/flat) — No raw payload exists for the prediction week.
59 - `2026-W23` `ShadowSpread/polymarket-auto-trading` (noise/flat) — No raw payload exists for the prediction week.
60 - `2026-W23` `pewdiepie-archdaemon/odysseus` (gap/up) — No raw payload exists for the prediction week.
61 - `2026-W23` `ssreeni1/tracebase` (gap/up) — No raw payload exists for the prediction week.
62 - `2026-W23` `Aimer-zero/redforge-ai` (gap/up) — No raw payload exists for the prediction week.
63
64 ## Editorial Readout
65
66 Signal calls are judged by later star growth, noise calls by limited follow-on growth or disappearance, and gap calls by the traction of the related edge repos cited in the blind-spot narrative. Gap scores are therefore weaker proxies than signal/noise scores and should guide reskill discussion rather than act as hard truth.