market-movers
Skill3 starsChanged 4 months agoMITView on GitHub
CoinGecko movers + categories brief with optional Kraken tradability cross-reference. Gainers, losers, trending (24h), and category rotation (CoinGecko /coins/categories) feed the morning brief. Each CoinGecko panel retries with exponential backoff (1s/2s/4s); final 429 on a panel degrades that panel with an explicit marker in `note`. Kraken tradability is sourced from `kraken pairs -o json` with an in-process TTL cache; CLI absence degrades the per-entry `tradable_on` field to `null` with an explicit marker.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- Kungfookenny123/aso-skillstheir own version, first seen Feb 27, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- Kungfookenny123/aso-skills★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository47 other files it gives its agents.
- AGENTS.mdAGENTS.md
- beadsSkill
- backtest-engineSkill
- backtest-pipelineSkill
- backtest-trend-minerSkill
- bug-scanSkill
- daily-trade-pickSkill
- execution-kraken-perpsSkill
- execution-kraken-spotSkill
- l3-conviction-scanSkill
- market-accumulationSkill
- market-basisSkill
- market-breadthSkill
- market-breakoutSkill
- market-emaSkill
- market-exhaustionSkill
- market-fibonacciSkill
- market-liquidity-sweepSkill
- market-macdSkill
- market-macroSkill
- market-notesSkill
- market-overviewSkill
- market-rsiSkill
- market-snapshotSkill
- market-squeezeSkill
- market-s-rSkill
- market-stateSkill
- market-trend-qualitySkill
- market-trendSkill
- market-valuationSkill
- market-volatilitySkill
- market-volumeSkill
- market-watchlistSkill
- portfolio-mgmtSkill
- position-watchdogSkill
- risk-engineSkill
- run-all-l2Skill
- run-all-l3Skill
- run-watchlistSkill
- strategy-accumulation-swingSkill
- strategy-breakout-confirmSkill
- strategy-exhaustion-fadeSkill
- strategy-fitness-heatmapSkill
- strategy-funding-carrySkill
- strategy-liquidity-sweepSkill
- strategy-mean-reversionSkill
- strategy-trend-followSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 27
When tradable_on is null (Kraken CLI missing/rate-limited) — do not treat a CoinGecko hot entry as "tradable on Kraken".
- Very specific: line 62
When enabled, kraken pairs -o json is invoked once per process and cached for --kraken-pairs-ttl-s seconds (default 600s).
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
market-movers
Cross-asset movers brief source. Pulls four panels for the morning brief, with one optional cross-reference against the user's execution venue:
- Gainers — top N by 24h % change (
/coins/markets?order=percent_change_24h_desc). - Losers — same payload, ascending.
- Trending — top N from
/search/trending(separate quota tier). - Categories — top N from
/coins/categoriesordered by 24h market-cap change (the "where is the rotation" signal next to per-coin movers). tradable_on— per-entry, optional. Cross-references each gainers/losers/trending entry againstkraken pairs -o jsonso the brief is actionable (you can actually execute it on Kraken), not just informative.
When NOT to use
- As a trade decision tool — this is a discovery/screening brief (what is hot), not a setup with entry/stop/conviction. Validate any name through
run-watchlist/run-all-l3before acting. - When
tradable_onisnull(Kraken CLI missing/rate-limited) — do not treat a CoinGecko hot entry as "tradable on Kraken"; the cross-ref signal is unavailable. - For execution — hand off to
execution-kraken-spot; this skill only reports, it never places orders.
Quick start
# JSON for LLM tool-use (preferred for batch + morning brief)
uv run skills/market-movers/scripts/run.py --json
# Human-readable
uv run skills/market-movers/scripts/run.py
# Wider panels / more retries
uv run skills/market-movers/scripts/run.py --top-n=10 --retries=5 --json
# CI / log-only run without the `kraken` CLI on PATH
uv run skills/market-movers/scripts/run.py --no-tradable-filter --jsonPanels + endpoint matrix
| Panel | Endpoint | Quota | Degrade when 429 |
| Gainers (24h top N) | /coins/markets?order=percent_change_24h_desc | shared ~30 req/min | drop + [MOVERS API RATE-LIMITED — gainers/losers unavailable this run] |
| Losers (24h top N) | same payload, sort ascending | same | same |
| Trending | /search/trending | higher tier | drop silently (rare) |
| Categories (24h rotation) | /coins/categories?order=market_cap_change_24h_desc | same bucket as /coins/markets | drop + [MOVERS API RATE-LIMITED — categories unavailable this run] |
| Tradable cross-ref | kraken pairs -o json (in-process TTL cache) | local CLI | empty fields + [MOVERS KRAKEN CLI UNAVAILABLE — tradable_on field empty this run] |
Each CoinGecko panel retries with exponential backoff (1s, 2s, 4s, ...)
up to --retries (default 3 → two backoff sleeps of 1s and 2s).
Trending is preserved when gainers/losers rate-limit; categories is a
separate panel so it can degrade independently.
The tradability cross-reference is opt-in via --no-tradable-filter.
When enabled, kraken pairs -o json is invoked once per process and
cached for --kraken-pairs-ttl-s seconds (default 600s). The CLI is
invoked directly via subprocess.run — no Python adapter layer.
CLI flags
| Flag | Default | Notes |
--top-n | 7 | Entries per coin panel (gainers/losers/trending). Brief uses 7; widen for a deeper dive. |
--categories-top-n | 10 | Categories panel size. 0 to skip the panel entirely. |
--retries | 3 | Per-panel retry count. Final 429 after N attempts → rate-limit escalation for that panel. |
--no-tradable-filter | off | Skip the kraken pairs lookup. Use in CI without the kraken CLI on PATH, or when brief size matters and the cross-ref isn't needed. |
--kraken-pairs-ttl-s | 600 | In-process cache TTL for the AssetPairs response. A daily brief makes at most one subprocess call per day at this default. |
--json | off | Emit JSON to stdout (LLM tool-use). |
Exit code is 0 on success and degrade alike — a rate-limited or
CLI-missing run is a known-degraded outcome, not an error. Scheduled
runs can disable the rate-limit log line via
MARKET_SKILLS_NO_RATE_LIMIT_LOG=1 if needed.
Output shape (JSON)
{
"fetched_at": "2026-06-29T05:00:00+00:00",
"gainers": [
{
"id": "bitcoin", "symbol": "BTC", "name": "Bitcoin",
"pct_24h": 1.42, "price_usd": 62150.0,
"market_cap_rank": 1,
"tradable_on": {"kraken": true, "altname": "XBTUSD",
"base": "XXBT", "quote": "ZUSD"}
}
],
"losers": [/* same shape, negative pct */],
"trending": [{"id": "...", "symbol": "...", "name": "...",
"market_cap_rank": 7,
"tradable_on": {"kraken": false}}],
"categories": [
{"id": "ai", "name": "Artificial Intelligence",
"market_cap_usd": 12300000000, "pct_24h": 12.34,
"top_3_coins_id": ["render-token", "fetch-ai", "the-graph"]}
],
"rate_limited": false,
"attempts": {"gainers_losers": 1, "trending": 1, "categories": 1},
"note": "",
"kraken_cli_available": true,
"tradable_filter": true
}tradable_on field
Three states per entry, depending on the tradability state at fetch time:
| Value | Meaning |
null | The Kraken AssetPairs cross-ref was attempted but the kraken CLI is missing (FileNotFoundError) or returned an unparseable payload. Read top-level kraken_cli_available to distinguish. The LLM agent brain should NOT treat this as "CoinGecko says this is hot AND I can trade it on Kraken" — the cross-ref signal is silently unavailable. |
{"kraken": false} | AssetPairs response parsed cleanly, no Kraken pair matched the entry. Often: low-cap coins not yet listed on Kraken, or symbol-collision edge cases. |
{"kraken": true, "altname", "base", "quote"} | The entry is tradable on Kraken. altname is the Kraken instrument name (e.g. XBTUSD) suitable for execution-kraken-spot submit. Quote preference is stablequote (ZUSD/USDT/USDC) > ZEUR/ZGBP > XBT > XETH. |
Symbol-collision caveat: a CoinGecko entry whose id is not in the
built-in _COINGECKO_ID_TO_KRAKEN_BASE map (top-50 by market cap)
falls back to symbol-only lookup, which can match a Kraken pair for a
different coin sharing the symbol (e.g. a low-cap chain that publishes
a token literally named BTC). The built-in map prevents the most
likely collisions; consumers that need a precise answer for an unmapped
coin should cross-reference the CoinGecko id against Kraken's own
listing pages.
note markers
Multiple degrade markers can stack. Examples:
- Gainers/losers rate-limited, Kraken CLI OK:
"[MOVERS API RATE-LIMITED — gainers/losers unavailable this run]" - Categories rate-limited only:
"[MOVERS API RATE-LIMITED — categories unavailable this run]" - Kraken CLI missing:
"[MOVERS KRAKEN CLI UNAVAILABLE — tradable_on field empty this run]" - Gainers/losers rate-limited AND Kraken CLI missing:
"[MOVERS API RATE-LIMITED — gainers/losers unavailable this run] [MOVERS KRAKEN CLI UNAVAILABLE — tradable_on field empty this run]"
The morning-brief prompt reads note and surfaces it verbatim so the
user knows the brief was partial.
Rate-limit log
Path: $XDG_DATA_HOME/market-skills/coingecko-rate-limit.log (one JSON
line per CoinGecko 429 incident — gainers/losers, trending, and
categories each log independently). The Kraken CLI lookup has no
rate-limit log; its absence degrades to tradable_on: null and is
captured in note. When XDG_DATA_HOME is unset, the CoinGecko log is
silently skipped (no console fallback). One record per panel —
endpoint is the full URL, attempts is the actual retry count,
final_status is the HTTP code that broke the retry chain.
Disable for clean test runs: export MARKET_SKILLS_NO_RATE_LIMIT_LOG=1.
Failure modes
- CoinGecko 429 × retries on
/coins/markets→ degrade that panel with the explicitnotemarker; trending stays populated. - CoinGecko 429 × retries on
/coins/categories→ degrade that panel with the explicitnotemarker; movers unaffected. krakenCLI absent → every entry'stradable_onisnull;kraken_cli_available: false;notecarries the explicitKRAKEN CLI UNAVAILABLEmarker.kraken pairs -o jsontimeout (default 30s) → same shape as CLI absent (caller can't distinguish).kraken pairs -o jsonreturns unparseable JSON → same shape as CLI absent.kraken pairs -o jsonreturns empty dict → empty index; entries gettradable_on: {"kraken": false}(legitimate "Kraken lists no online pairs" answer).- Network error × retries on CoinGecko → empty payload,
rate_limited: false(we don't conflate network errors with throttling). - Schema change (CoinGecko renames a field) → field is
Nonein the output; no crash, no alert. Consumers that need a specific field should treatNoneas "unavailable". - Empty response → empty panel, no note.
Data flow
lib.fetch_movers(*, top_n, retries, tradable_filter, categories_top_n,
kraken_pairs_ttl_s, sleeper, kraken_runner, now_s) is a pure-function-
shaped call. sleeper defaults to time.sleep and tests pass a no-op;
kraken_runner defaults to subprocess.run and tests pass a stub;
now_s defaults to time.time and tests pass a fixed-time callback.
The CLI parses flags, calls the lib, prints either JSON or a formatted table. No DB, no portfolio context, no env-mutating side effects beyond the CoinGecko rate-limit log.
Output envelope (AXI)
--json output follows the canonical AXI envelope — {data, count, errors, help[]}. Default schema is the per-skill minimal fields (3-6 essentials); pass --fields=<csv> to project or --full for the full payload. count is the item count, help[] is contextual next-step command templates. Lib.py return shapes (L1Result / L2Result / L3Result / L3Idea / RegimeSignal) are unchanged — the envelope wraps them at the scripts/run.py boundary.
Home view (no-arg mode)
No-arg mode prints the home view from $XDG_DATA_HOME/market-skills/<skill>_last.json (last successful run). Errors are not cached.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

