Unlock Monitor

<!-- autoresearch: variation B — sharper output via Absorption Ratio (unlock $ / avg daily volume), Cliff vs Linear classification, and a one-line market-read verdict per unlock. Replaces qualitative HIGH/MED/LOW tiers with quantitative liquidity-strain thresholds backed by Keyrock's 16k-unlock study. Folds in source-status observability (from C) and CoinGecko volume enrichment (from A) as cheap wins. -->

Read memory/MEMORY.md for context on market positions and active narratives. Read the last 7 days of memory/logs/ to dedup against any unlock already covered. Read memory/state/unlock-monitor-seen.json if present — a list of ${ticker}:${unlock_date} keys already shipped. Skip exact matches; if file is absent, treat as empty.

Core thesis

The original skill ranked unlocks by % of circulating supply. That's a weak proxy. The right metric is Absorption Ratio = unlock_dollar_value / 7-day avg daily volume. Keyrock's analysis of 16k+ unlocks shows that ratios above ~2.4× consistently strain liquidity and produce measurable price drawdown. Below ~0.5× the market typically yawns. This skill ranks by absorption ratio first, supply % second, and folds in cliff-vs-linear structure plus pre-unlock price action to produce a per-event verdict.

Steps

1. Gather candidate unlocks

Run these WebSearches in parallel:

• "token unlock schedule" "${week_of}" site:tokenomist.ai OR site:defillama.com
• "token unlock" "this week" cliff vesting team investor ${current_year}
• "upcoming unlocks" cryptorank OR dropstab ${current_year}
• "FTX distribution" OR "Mt. Gox" OR "Celsius" OR "creditor payout" crypto ${current_year} (court-ordered supply shocks count)
• "$10M" OR "$50M" OR "$100M" token unlock ${week_of} (dollar-value angle catches what supply % misses)

Also fetch (WebFetch fallback if any URL fails):

• https://tokenomist.ai — primary, source-verified across 500+ tokens with cliff/linear labeling
• https://defillama.com/unlocks — DeFi protocols, with $ values
• https://cryptorank.io/token-unlock — broad coverage, recipient-category labeled
• https://dropstab.com/vesting — alternative cross-source verification
• https://www.coingecko.com/en/highlights/incoming-token-unlocks — Tokenomist-powered

Record which sources returned data → emit a sources: line in the log (see step 7).

2. Enrich each candidate with volume + structure

For each candidate unlock, fetch from CoinGecko search results or WebFetch on https://www.coingecko.com/en/coins/${slug}:

• 7-day average daily volume (USD) — denominator for the absorption ratio
• Spot price at fetch time
• 30-day price change % — input to the market-read verdict
• Vesting structure: cliff (lump-sum unlock) vs linear (gradual) vs mixed — usually labeled by tokenomist / cryptorank
• Recipient category: team, investor, ecosystem, community, creditor

If volume data is unavailable, mark the unlock vol=unknown and tier it conservatively — do not skip it, but flag the gap in the output.

3. Compute Absorption Ratio and tier

For each unlock with known volume:

absorption_ratio = unlock_usd_value / avg_daily_volume_usd_7d

Tier the events:

Recipient override: a team or investor unlock with absorption ratio one tier lower than CRISIS gets bumped up one tier (cost-basis-zero sellers act differently than airdrop recipients). A community or staking-reward unlock with ratio one tier higher gets bumped down one tier.

Court-ordered distributions (FTX, Mt. Gox, Celsius) bypass the tier system — always include, label forced, note the legal timeline.

4. Classify pre-unlock market read

For each event, set a one-line market_read based on 30d price change and vesting type:

• priced in — token is down >20% over 30d AND ratio ≤ STRAIN. Selling has already happened; unlock day may mark a bottom.
• market asleep — token is flat or up over 30d AND tier is STRAIN or CRISIS. Asymmetric downside; the move hasn't started.
• fade pump — token is up >15% over 30d AND tier is CRISIS. Classic pre-cliff bid-then-dump pattern.
• forced sellers — court-ordered. Different beast — legal timeline, not market-driven.
• absorbable — TRIVIAL or DIGESTIBLE with no recipient flag. Nothing to see here.

For cliff unlocks: weakness usually starts ~30 days before; vol peaks at unlock; recovery 10–14 days later. Note this pattern explicitly when relevant.