Kalshi Weather Markets

What it does

When your agent uses it

• Business, Finance & HR work in your project

Example prompts

• “/kalshi-weather-markets”

Requirements

• Python 3

Workflow steps

10 steps, taken from the first numbered list in SKILL.md.

1. 1Wrong settlement source. Scoring against a derived truth that correlates with but differs from the venue's resolution flips ~10% of…
2. 2Bracket off-by-one (phantom +1640%). Treating 2°F inclusive brackets {floor, cap} as 1°F half-open [floor, cap) manufactures a large…
3. 3strike_type not inferable from ticker. T74 on a low market might be greater or less. Always read strike_type from the API. Never guess.
4. 4Date-in-ticker, not close_time. Use the date embedded in the ticker string for settlement-date joins, not close_time (which is next-day…
5. 5LST ≠ local clock. Kalshi settles on LST (no DST). In spring/fall, the LST window shifts relative to local time. Cross-referencing WU…
6. 6CLI ≠ METAR. Raw ASOS hourly max/min is not the settlement value. CLI applies QC, backup-station fallback, and LST aggregation. Fit…
7. 7UTC vs local-day feature aggregation. Aggregating forecast features over UTC days instead of LST days misaligns labels — cost ~14…
8. 8Clock-mismatch look-ahead. Filling at an 18:00Z book snapshot while features are cut at 14:00 LST trades non-Eastern cities on future…
9. 9Phantom penny asks. 1¢ ask levels are frequently spoofed; assuming you fill them over-credits PnL ~23×. Count only depth that persists…
10. 10Overround as a diagnostic. Sum the YES prices across an event's full bracket set. overround > 1.0 is normal (house edge); overround >> 1.1…

What it can do on your machine

Read from SKILL.md and the folder at commit 981e1d7. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships 1 file in scripts/ (Python), which the agent can run.

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  Links to these hosts (documentation or services it may open):

  • docs.kalshi.com
  • weather.gov
  • mesonet.agron.iastate.edu
  • wunderground.com
  • docs.uma.xyz

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Kalshi Weather Markets loads about 2.2k tokens when it runs, and up to ~4.5k if it reads all its reference files. Until then it costs about 52 tokens; SKILL.md has 1,001 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~52

When it runs · the whole SKILL.md, loaded when a task matches

~2.2k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~4.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from agiprolabs/claude-trading-skills at commit 981e1d7, republished under its MIT licence (© agiprolabs). 1,001 words, ~2,196 tokens.

name

kalshi-weather-markets

description

Daily temperature high/low bracket and threshold contracts on Kalshi — contract structure, forecast→P(YES) map, settlement rules, cross-venue divergences, and weather-specific pitfalls

KXHIGH<CITY>-<YYMONDD>-B<center>     # bracket high
KXLOW<CITY>-<YYMONDD>-T<strike>      # threshold low

# Bracket B<center>, covering integers {floor, cap}
P(YES) = Φ((cap + 0.5 − μ) / σ) − Φ((floor − 0.5 − μ) / σ)

# Threshold "greater":
P(YES) = 1 − Φ((T + 0.5 − μ) / σ)

# Threshold "less":
P(YES) =     Φ((T − 0.5 − μ) / σ)

Φ(x) = 0.5 · (1 + erf(x / √2))   # stdlib only, no scipy needed

sigma_raw = max((p90 − p10) / 2.56, 0.5) · sigma_scale · sigma_mult
mu        = p50                          # or nowcast-blended (see forecasting.md)
sigma     = max(sigma_raw, 0.1)          # hard floor against degeneracy

mu_cli = mu_metar + bias_city_season     # bias = oracle_extreme − asos_extreme, fit per city + season

© agiprolabs, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 3 other files (scripts, references) in skills/kalshi-weather-markets of agiprolabs/claude-trading-skills.

• SKILL.md
• references/brackets-and-settlement.md
• references/forecasting.md
• scripts/weather_brackets.py

Open the folder on GitHubat commit 981e1d7